Retrospective Cohort Study Comparing Antiretroviral Treatment Outcomes Among Adolescents in Teen Clubs and Standard Care Clinics: Blantyre,Malawi

Abstract

Introduction

This study investigated if the Teen Club model improves virological suppression and reduces virological failure. Viral load monitoring is a golden ART programme performance indicator. HIV treatment outcomes are poor among adolescents compared to adults. Different service delivery models are being implemented to address this; among them is the Teen Club model. Currently, teen clubs improve treatment adherence (short-term impact), but there is a knowledge gap regarding the long-term impact. The study compared the rate of virological suppression and failure among adolescents in Teen Clubs and those on the standard of care (SoC) model.

Methods

This was a retrospective cohort study. A total of 110 adolescents in teen clubs and 123 adolescents in SOC from six health facilities were selected using stratified simple random sampling. The participants were followed for 24 months. STATA version 16.0 was used for data analysis. Univariate analyses were performed for both demographic and clinical variables. A Chi-squared test was used to assess the differences between proportions. Crude and adjusted relative risks were calculated using a binomial regression model.

Results

At 24 months, 56% of adolescents in the SoC arm had viral load suppression compared to 90% in the Teen Club arm. Of those who achieved viral load suppression at 24 months, about 22.7% (SoC) and 76.4% (Teen Club) achieved undetectable viral load suppression rates. Adolescents in the Teen Club arm had a lower viral load than those in the SoC arm (adjusted RR 0.23, 95% CI: 0.11-0.61; p = 0.002 adjusted for age and gender). Teen Club and SoC adolescents had virological failure rates of 3.1% and 10.9%, respectively. The adjusted RR was 0.16, 95% CI: 0.03-0.78; p = 0.023; those in Teen Clubs were less likely to have virological failure relative to those in SoC after adjusting for age, sex, and place of residence.

Conclusion

The study found that Teen Club models are more effective at achieving virological suppression among HIV positive adolescents.

Keywords

HIV adolescents teen club antiretroviral therapy viral load

Introduction

In 2019, approximately 1.7 million adolescents (10-19 years) globally were living with HIV, 88% being in Sub-Saharan Africa.¹ In Africa, AIDS is the leading cause of death among adolescents and second highest globally.² In 2019, about 34,000 adolescents died of AIDS-related illnesses. Malawi estimated AIDS-related deaths in 2020 were 13,100.² Despite the massive rollout of antiretroviral therapy (ART) for decades, adolescent still have poorer treatment outcomes at all stages of the HIV continuum, including viral load suppression.³ Numerous factors have been associated with this phenomenon, ranging from late initiation to poor adherence to treatment. Challenges include; social, health system related and treatment related.^4–6 The unique developmental, psychosocial, behavioural and infrastructural factors affect HIV treatment outcomes.^7,8 All this has made HIV management complex among adolescents.

After appreciating that ART services were not a one size fits all, differentiated service delivery models (DSDs) were introduced.⁹ The Teen Club model was designed to provide adolescent friendly ART clinics.^4,5 This DSD comprises of different services assumed to address the challenges being faced by adolescents on ART. Currently not all health facilities have Teen Clubs in Malawi.

The goal of ART is to suppress viral load hence improve quality of life for PLHIV. Malawi has adopted ART guidelines stipulating intervals (6, 12 and 24 months) for viral load routine monitoring after ART initiation.⁴ The guidelines have a provision for targeted monitoring for individuals with high viral load.

Currently, there is substantive evidence that the Teen Club models are better than Standard of Care in levels of adherence among adolescents on ART. Most of these studies were conducted soon after Teen Clubs were introduced hence focused on adherence, an immediate outcome of the program. However, there is little evidence highlighting the long-term impact of these clubs.

According to WHO, viral load monitoring is the gold standard approach for monitoring HIV treatment and it is a key performance ART program indicator.⁵ Only one study was done is Namibia,¹⁰ however, the standard of care was a specialised paediatric ART clinic⁶ which is different from Standard of Care clinic in Malawi.

The study was aimed at evaluating the impact of Teen Clubs among adolescents living with HIV on ART focusing on viral load suppression at 24 months and virological failure rates at 12 months. Filling this knowledge gap might reveal the need to scale-up Teen Club model.

Methods

Study Design

This was a 2-year retrospective cohort study. The study utilised data from both electronic medical records (EMR) and hard copies to follow-up clients. Most of the health facilities in Blantyre city have J2,⁷ an electronic data management device. A case report form was designed to extract data.

Study Population

The study population included ALHIV aged between 10 and 19 years initiated on ART in Blantyre within selected health facilities between 1^st of January 2018 to 31^st of December 2020. The adolescents had at least attended one ART clinic.

The Teen Club and Standard of Care Model

The Teen Club model was first piloted in Uganda in 2003.^8,11 The model aims at improving the quality of HIV care among ALHIV. The model aims at improving both clinical and mental outcomes by building positive relationship, improving self-esteem, acquiring life skills, adult role modelling and structural adolescent-focused adherence support activities.^11,12 Malawi ministry of health adopted and recommended the scale-up of the Teen Club model in 2007. The model is more comprehensive than the Standard of Care, see Table 1.^11,13

Table 1.

A Comparison of Teen Club and Standard of Care Models.

Model	Teen Club	Standard of Care
Services offered	Clinical review ART re-scripting ART refill Family planning Labs Referrals Structured adolescent-focused adherence support Psychosocial support	Clinical review ART re-scripting ART refill
Personal	Dedicated ART service provider with extra training in managing adolescent ART clinics. Social workers	ART service provider
Operating Time	Services provided outside normal working hours. Clinics are conducted during weekends, once or twice per month	Services provided during normal working hours
Location	Special room with adequate space outside for recreational activities	ART clinic used by all ART clients despite age
Population	Open for adolescent only	Mixed population

Inclusion and Exclusion Criteria

Those HIV positive, initiated on ART between January 2018 and December 2020, aged between 10 and 19 years and residence of Blantyre were eligible for the study. The exclusion criteria was as follows; those above 19 years, adolescents’ imitated on ART before 2018 or after 2020 and all adolescents transferred in from other facilities.

Variables of Interest

Below are variables of interest for this study

Viral Load Suppression

Viral load is defined as the amount of HIV in a sample of blood and is reported as the number of HIV RNA copies per millilitre of blood.¹⁴ Results can be classified into 3, <40 copies/ml is undetectable viral load; 40-999 copies per ml is low viral load and >999 copies/ml is high viral load. The primary classification of the study LDL was defined as <999 copies/ml and HDL being defined as >999 copies/ml.

Virological Failure

This is defined as a type of treatment failure that occurs when ART fails to suppress and sustain an individual's viral load to <200 copies/ml of blood.¹⁴ This was tested between 12 and 18 months after ART initiation. Those with HDL were subjected to intensive adherence counselling and re-tested for VL. Those whose results were still >200 were taken as suspects of virological failure.

Other Variables of Interest

The other variables collected and used in the study included; age, sex, area of residence (urban/rural), reason for linkage to care, individual's duration on ART, type of model in (Teen Club/Standard of Care), HIV treatment line and history of opportunistic infections especially TB.

Sample Size and Sampling Techniques

The sample size was calculated based on the estimated viral load suppression rates at 24 months between adolescents in Teen Clubs and those in Standard of Care. Adherence rate was used as a proxy to treatment success (viral load suppression). Differences in adherence rates between the two arms were obtained from a study done in Malawi.¹⁵ Adherence rates for adolescents in Teen Club and Standard of Care were 77.2% and 49.5%, respectively. Power was at 90%, 95% confidence interval, and the difference was estimated at 27.7% between the two arms. An estimated 10% dropout rate was included in the sample size. The minimal sample size was calculated using the following formula.

n = \frac{{z_{1 - (a / 2)} + z_{1 - β}}^{2} [p_{1} (1 - p_{1}) + p_{2} (1 - p_{2})]}{{(p_{1} - p_{2})}^{2}}

The minimum calculated sample was 128 participants for the two groups (64 per group). The study sample was all-inclusive, and a total number of 233 study participants were recruited. A total of 123 from Standard of Care and 110 from the Teen Club arm. More participants were recruited because not everyone had an opportunity to be tested for viral load suppression at 24 months.

Health facilities were stratification by the model of service delivery provided. Twenty-two health facilities provided Teen Club services while 9 health facilities provided Standard of Care.

Simple random sampling was used to identify final facilities. The study sample was all-inclusive of the study population. A list of eligible participants was obtained through medical records.

Data Collection

After obtaining ethical clearance from the College of Medicine Research and Ethics Committee (COMREC), the Blantyre District Health Office authorised the study. Officers in charge for study sites were notified about the study and their role. Two nurses were recruited and oriented as data collectors.

Data Management and Analysis

Data were collected using a standardised pre-tested CRF. Data elements were extracted from EMR, hard copy registers, and viral load results slip. In facilities using J2, hard copies were still used to check for missing data. Each questionnaire had a unique identity for easy tracking. Data were then entered into Microsoft excel for cleaning and storing. The excel file was exported to STATA, and data was coded.

STATA version 16.0 was used for data analysis. Univariate analysis was performed to describe both demographic and clinical variables. A Chi-squared test was used to assess the difference between proportions of virological suppression and virological failure rates in the two study arms. The cut-off point for significance was p < 0.05. Relative risk was calculated using a binomial regression for comparison of virological suppression and failure levels between the two arms.

Bivariate analysis was performed to test for association between age groups, sex, and residence against both viral load suppression and viral load failure. Adjusted RR for both virological suppression and failure were calculated using a multivariate model.

Data Presentation

Descriptive data were presented using tables and charts. Analytical data showing differences between the models were presented using proportions, crude relative risk, and adjusted relative risks with 95% confidence intervals and p-values.

Ethical Approval and Informed Consent

This study was approved by the College of Medicine Research Ethics Committee (CoMREC) and the reference number is Alibi3308. Approval was also obtained from Blantyre District Health Office. No informed consent was sort from participants as the study relayed on medical records only. The researcher had no direct contact with study participants. The case report forms (CRF) used unique identity numbers to identify participants making them anonymous.

Results

Characteristics of Study Participants

A total number of 289 ALHIV attended ART clinics in the six study sites during the study period. About 53% of them were in the Standard of Care arm. Only 17% of those in standard of care were registered in health facilities with Teen Clubs.

The mean age of participants was 15 years with a standard deviation of about 3.4. While mode age was 16 years with an interquartile range of (13-19 years). Most of the adolescents (62%) were within the older category (15-19 years).

About 62% of the participants were linked to care through the test and treat protocol and 23% were linked through the PMTCT program. Of those linked through the PMTCT program only 2% belonged to the Teen Club arm.

The mean time on ART among the participants was 28 months with a standard variation of 6 months. At 24 months, the percentage of adolescents retained in Teen clubs and Standard of Care was 90% and 73%, respectively. About 80% of defaulters were in the Standard of Care arm. Table 2 summarises the difference in demographic variables.

Table 2.

Demographic and Clinical Variables for Study Participants.

Characteristics	Total	Standard care n (%)	Teen club n (%)	p-Value
Sex
Male	70	26 (21)	44 (40)	0.002*
Female	163	97 (79)	66 (60)
Age group
10-14 years	89	34 (28)	55 (50)	0.001*
15-19 years	144	89 (72)	55 (50)
Residence
Rural	102	102 (83)	0 (0)	0.001*
Urban	131	21 (17)	110 (100)
Reason for linkage
Test and treat	144	75 (61)	69 (63)	0.001*
Advanced disease stage	60	21 (17)	39 (35)
PMTCT	29	27 (22)	2 (2)
Opportunistic infections
Yes	20	6 (5)	14 (13)	0.034*
No	212	117 (95)	96 (87)
Time of ART (months)
0-23 months	53	23 (19)	30 (27)	0.12
25+ months	180	100 (81)	80 (73)

*Association is significant at 0.05 level.

Virological Suppression Rates

The overall viral load suppression rate at 24 months was 73%. About 17% of the cohort was not monitored for viral load suppression at 24 months with the majority (78%) being from the Standard of Care arm. Adolescents in Teen Clubs had higher viral load suppression rates as compared to those in the Standard of Care arm, and the difference was significant at all levels of assessment (6, 12, and 24 months). About 90% of adolescents in Teen Clubs achieved viral load suppression at 24 months compared to 56% for the adolescents in the Standard of Care arm (p = 0.001). Of those who achieved viral load suppression at 24 months, about 22.7% (Standard of Care) and 76.4% (Teen Club) achieved undetectable viral load.

Table 3 describes and compares the proportion of viral load suppression between adolescents in the study arms. Adolescents in Teen Club had a relatively lower risk of having higher viral load at 6, 12, and 24 months compared to those in the Standard of Care arm (p = 0.001 at all levels). After controlling for age, sex and residence the adjusted risk ratio remained significant (p = 0.002), with adolescents in Teen Clubs having a 74% relatively lower chance of having high viral load suppression rates at 24 months as compared to those in Standard of Care. After testing for possible predictors, only area of residence was significant but in the multivariable model, all factors (age, sex, and residence) were insignificant at 24 months.

Table 3.

Differences in Viral Load Suppression Rates.

	6 months	12 months	24 months
Characteristics	Suppressed n (%)	Suppressed n (%)	Suppressed n (%)	Crude risk ratio (95% CI)	Adjusted RR (95% CI)
Model of care
Teen club	96 (88)*	93 (91)*	72 (90)*	0.23* (0.11-0.46)	0.28* (0.11-0.61)
Standard care	66 (65)	41 (67%)	44 (56)	1	1
Age group
15-19 years	103 (82)*	76 (80)	74 (74)	0.9 (0.54-1.51)	0.67 (0.42-1.1)
10-14 years	59 (69)	58 (85%)	42 (71)	1	1
Sex
Female	112 (78)	87 (78)	79 (70)	1.5 (0.8-2.9)	1.4 (0.74-2.5)
Male	50 (75)	47 (90)	37 (80)	1	1
Residence
Urban	111 (87)*	103 (90)*	84 (85)*	0.33* (0.19-0.56)	0.8 (0.44-1.5)
Rural	51 (61)	31 (65)	32 (53)	1	1

*Association is significant at 0.05 level.

Furthermore, at 6 and 12 months, RR 0.34, 95% CI: 0.2-0.6; p = 0.001 and RR 0.21, 95%CI: 0.13-0.55; p = 0.001, respectively. Adolescents in Teen Clubs had a relatively lower risk of having high viral load as compared to those in the Standard of Care arm.

However, age was only significant at 6 months, showing that older teens (15-19 years) had a higher viral load suppression rate compared to younger adolescents (p = 0.04). At 6 months after ART initiation, older adolescents had a 40% relatively lower chance of having high viral load compared to younger adolescents RR = 0.6, 95% CI: 0.37-0.97; p = 0.039. At 12 and 24 months, age ceased to be a significant factor among adolescents on ART.

At 6-, 12-, and 24-month adolescents living in urban areas were less likely to have a high viral load as compared to those in rural areas (p = 0.001 at all levels) and in both arms. However, in the adjusted model, area of residence seised being significant (p = 0.32). The overall proportion of adolescents with viral load suppression among the adolescents living in urban areas was higher as compared to those living in rural areas, 86% versus 53%. Adolescents in urban areas were 68% relatively less likely to have high viral load compared to those living in rural areas, RR 0.32, 95%CI: 0.19-0.56; p = 0.001. But after adjusting for all variables, area of residence seised to be significant.

Figure 1 below is a graph showing viral load monitoring results for both arms at 6, 12, and 24 months. The results show a higher proportion of adolescents with undetectable viral load (>40 copies/ml) in Teen Clubs as compared to the Standard of Care arm. Furthermore, the graph shows that the Standard of Care arm had the largest proportion of adolescents with LDL (40-999 copies/ml) as compared to those in the Teen Club arm.

Figure 1.

A comparison of viral load results between the two arms at 6, 12 and 24 months.

Virological Failure Rates

The overall virological failure rate among adolescents monitored for viral load suppression at 12 months was 6%. Table 4 below shows a comparison of virological failure rates and risk ratios for adolescents in the two study arms. The proportion of virological failure between participants in Teen Clubs and those in Standard of Care were 3.1% and 10.9%, respectively, although the difference was not statistically significant. After adjusting the relative risk for age, sex and area of residence, the model showed the likelihood of developing virological failure among adolescents in Teen Club was relatively reduced by 84% compared to those in the Standard of Care (p = 0.023).

Table 4.

Virological Failure Rates and Risk Ratios for Adolescents in the Two Arms.

Characteristics	VL failure proportion n (%)	Crude risk ratio (95% CI)	Adjusted RR (95% CI)
Mode of care
Teen club	3 (3.1)	0.29 (0.07-1.15)	0.16 (0.03-0.78)*
Standard care	5 (10.9)	1	1
Age group
15-19 years	2 (2.5)	0.27 (0.06-1.27)	0.23 (0.05-1.11)
10-14 years	6 (9.5)	1	1
Sex
Female	5 (5.4)	0.91 (0.23-3.63)	0.95 (0.25-3.6)
Male	3 (6)	1	1
Residence
Urban	5 (4.6)	0.52 (0.13-2.1)	1.9 (0.38-9.0)
Rural	3 (8.8)	1	1

*Association is significant at 0.05 level.

Discussion

The general virological suppression rate at 24 months was below the 2020 UNAIDS HIV treatment target third 90.¹⁶ This justifies the need for prioritising adolescents if Malawi is to attain the third 95 on the 2025 UNAIDS treatment targets. Furthermore, the overall virological suppression rate was slightly lower than that of the study done in KwaZulu Natal, South Africa (81%).¹⁷ As compared with adults, this shows that adolescent HIV treatment outcomes are still very poor in Malawi, however this is a common trend in most countries in Sub-Saharan countries including South Africa.^10,18

Having more than a 30% difference in the virological suppression rate shows that interventions done in Teen Club effectively contribute to HIV treatment success. Among other activities being implemented at the Teen Club are treatment literacy, adherence and mental health.^13,19 These areas are key, especially mental health, for the adolescents living with HIV as they face unique challenges ranging from mental, structural, social and behavioural,^3,12 Furthermore, having about 80% of those with undetectable viral load from the Teen Club arm shows how effective the program is in improving treatment outcomes among adolescents. This also implies that the Teen Club model can effectively contribute towards reducing HIV infection among adolescents to zero. This might be the reason why UNICEF and USAID documented Teen Clubs as a best practice and promising practice respectively towards improving adolescent HIV treatment outcomes.¹⁹

The viral suppression rate among adolescents in Teen Clubs was 90% compared with 56% in the Standard of Care arm. The results were similar to the study done in South Africa¹¹ with 91% versus 80% in adolescent HIV friendly clinics and Standard of Care arm, respectively.

However, the study findings differed from the findings of the study done in Namibia^6,10 that revealed that there was no significant difference between the two arms. The difference in the study results was due to the difference between the comparison groups. In Namibia, the Standard of Care was a specialised paediatric HIV clinic with service providers trained in adolescent HIV treatment and providing extra services such as adherence counselling and treatment literacy education.

Moreover, the study also revealed that there is a significantly high proportion of adolescents with low viral load in urban areas as compared to rural areas. Furthermore, adolescents in urban areas were less likely to have a high viral load at 24 months compared to those in rural areas. The findings are similar to the study findings on the vulnerability of children on ART in rural public health centres conducted in South Africa.^20,21 That study revealed that in rural health centres, ART outcomes were poor because of weak health systems which might also be a contributing factor in Malawi.

The overall virological failure rate for the study participants was 6% at 12 months which was almost the same as that of the 2020 general population in Malawi.¹⁰ Viral load failure rate was not significantly different between the adolescents in Teen Club and those in Standard of Care (3% and 11%, respectively). After adjusting for age and sex, the results showed that those in Teen Clubs have an 84% relatively lower chance of having virological failure as compared to adolescents in the Standard of Care arm.

The study further revealed that there is no significant difference between the younger and older adolescents in virological failure proportions. These findings differ from the South Africa, Mpumalanga:^14,22 that reported younger adolescents are more at risk of virological failure compared to older adolescents. The difference between the two studies might be due to the difference in the cut-off points for virological failure. The Mpumalanga study had a higher cut-off point of >400 copies/ml different from this study.

Despite having 6% of the participants at 12 months showing signs of treatment failure, all of them remained on first-line ART regimen. During the 2-year follow-up period, no participants were switched to second line, and no one was referred for genotyping.

The study had some challenges as other study sites had no EMR although hardcopy registers had to be used to extract data. It was difficult for the team to establish history of OIs for the study participants in the two sites without EMR. Some of the EMR were incomplete, especially viral load records were not entered and the team had to enter the records. Again, there was inconsistence in recording of viral load data within and between health facilities. Values and categories were being used interchangeable and sometimes with different cut-off points. Finally, some of the facilities entered wrong dates of birth for the clients and the team had to double check using the master cards.

Involving the ART/PMTCT coordinator was a strength, as most of the challenges faced during data collection were addressed on spot. This resulted in the team having clean and complete dataset for analysis.

Conclusion

The study results show that the Teen Club model is more effective towards attaining virological suppression among adolescents living with HIV. If the country wants to ensure adolescents do not only reach viral load suppression but also undetectable levels, then there is a need to increase the coverage of Teen Clubs. The study also showed that adolescents in Teen Club are less likely to have virological failure as compared to those in the Standard of Care arm. For the country to attain the third 95 UNAIDS treatment target by 2030 there is an urgent need to scale-up the Teen Club model.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Michael Alibi

References

UNICEF. HIV and AIDS in adolescents [Internet]. [cited 2020 Jul 17]. Available from: https://data.unicef.org/topic/adolescents/hiv-aids/, 2020

UNAIDS. UNAIDS 2016-2021 Strategy—On the Fast-Track to end AIDS. UNAIDS; 2016:20–30.

Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in Adults and Adolescents with HIV. Department of Health and Human Services; 2021:40–51.

CDC. Guidelines for the use of ARV Agents in Adults and Adolescents Living with HIV. CDC; 2019:9–16.

Wyk

Davids

L-AC

. Challenges to HIV treatment adherence amongst adolescents in a low socio-economic setting in Cape Town. S Afr J HIV Med. 2019;20(1):a1068.

Management Sciences for Health. In: Mpawa

Mchacha

INC

, eds. District Health System Strengthening and quality improvement for service delivery retain: Teen Clubs help retain adolescents in HIV care and treatment. Management Sciences for Health; 2018:1–4.

Lewis

Abramowitz

Koenig

Chandwani

Orban

. Negative life events and depression in adolescents with HIV: a stress and copying analysis, AIDS Care. AIDS Care [Internet]. 2015; Available from: https://pubmed.ncbi.nlm.nih.gov/26313848/.

ICAP. In: West

Schoeneborn

, eds. ICAP approach to differentiated service delivery. Columbia University, Mailman School of Public Health; 2015:15–20.

International AIDS Education and Training Centre. Implementation of HIV/AIDS DSD models and healthcare process outcomes in Kenya. Siaya; 2012.

10.

Munyayi FK, van Wyk BE. The comparison of teen clubs vs. standard care on treatment outcomes for adolescents on antiretroviral therapy in Windhoek, Namibia. AIDS Res Treat. 2020;2020:8604276. doi:10.1155/2020/8604276.

11.

Baylor College of Medicine. Malawi Teen club curriculum part 1: content A resource for groups working with adolescents living with HIV. Baylor College of Medicine; 2012:5–26.

12.

Ministry of Health Malawi. Adolescent and Psychosocial Services : BIPAI Teen Club Model – Malawi Differentiated Care Model for Malawi Overview. Lilongwe; 2020.

13.

Baylor College of Medicine. BIPAI Teen Club. Available from: https://www.texaschildrensglobalhealth.org/bipai teenclub#.

14.

U.S Department of Health and Human Services. Glocery of HIV/AIDS-Related Terms. 9th ed. USA; 2021.

15.

Mackenzie

Lettow

Gondwe

, et al. Greater retention in care among adolescents on antiretroviral treatment accessing “Teen Club” an adolescent-centred differentiated care model compared with standard of care : a nested case – control study at a tertiary referral hospital in Malawi. 2017.

16.

Malawi Ministry of Health. Malawi guidelines for clinical management of HIV in children and adults. 3rd ed. Ministry of Health; 2016:19–69.

17.

Zanoni

Sibaya

Cairns

Lammert

Haberer

. Higher retention and viral suppression with adolescent-focused HIV clinic in South Africa. PLoS One [Internet]. 2017;12(12):1–12. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0190260. Available from.

18.

WHO. The use of antiretroviral drugs for treating and preventing HIV infection: recommendations for a public health approach. 2nd ed. WHO; 2016.

19.

Baobab Health Trust. Key program component [Internet]. Baobab Health Trust. 2008 [cited 2022 May 8]. Available from: https://healthmarketinnovations.org/program/baobab-health-trust

20.

Maskew

Fox

Evans

, et al. Insights into Adherence among a Cohort of Adolescents Aged 12–20 Years in South Africa : Reported Barriers to Antiretroviral Treatment. 2016;2016.

21.

Fatti

Bock

Grimwood

Eley

. Increased vulnerability of rural children on antiretroviral therapy attending public health facilities in South Africa : A retrospective cohort study. J Int AIDS Soc [Internet]. 2010;13(1):46. http://www.jiasociety.org/content/13/1/46. Available from.

22.

Evans

Menezes

Mahomed

, et al. Treatment outcomes of HIV-infected adolescents. 2013;29(6):892–900.