Past and current status of adolescents living with HIV in South Africa, 2005–2017

Objectives This paper reports HIV prevalence, incidence, progress towards the UNAIDS (90-90-90) targets, and HIV drug resistance among adolescents living with HIV in South Africa. We conducted secondary analyses using data extracted from the South African national HIV prevalence surveys (2005–2017). Analyses were stratified by sex and age (10–14 and 15–19-years), presenting weighted descriptive statistics, and realised totals. Results HIV prevalence increased from 3.0% in 2012 to 3.7% in 2017, translating to 360 582 (95% CI 302 021-419 144) HIV positive adolescents in 2017. Female adolescents bear a disproportionate HIV burden of 5.6% prevalence versus 0.7% for males. HIV incidence remained relatively stable. For the UNAIDS 90-90-90 targets, approximately 62.3% of adolescents knew their HIV status, 65.4% of whom were on antiretroviral therapy, and of these 78.1% on antiretroviral therapy had attained viral load suppression. There are knowledge gaps pertaining to the magnitude of perinatal infections and postnatal infections, and socio-behavioural risk factors for HIV transmission among adolescents in South Africa. There is still a need for focussed interventions targeting adolescent (1) gender disparities in HIV risk (2) screening for HIV, (3) sustained access and adherence to antiretroviral therapy and (3) retention in care to maintain viral load suppression.


Introduction
In 2020 approximately 1.7 million children under 15 years were living with HIV (ALHIV) globally; majority of whom reside in sub-Saharan Africa [1].Despite achievements in HIV containment strategies, South Africa, has the largest population of people living with HIV (PLHIV) estimated at 7.9 million, with adolescents most at risk [2,3].Adolescent HIV-related mortality remains high [4,5].Hence, prevention and early diagnosis of HIV infection among adolescents are critical [6][7][8].HIV infection in adolescents can be acquired postnatally and perinatally, the former due to documented structural and behavioral factors and the latter due to the differential coverage and uptake of mother-to-child transmission programmes [9][10][11].
Improved understanding of the circumstances of ALHIV is critical in designing adolescent-friendly health services to facilitate their successful transition into adult care.Compared to adults, adolescents have lower rates of HIV testing, disclosure [12], treatment adherence [13] long-term immunologic recovery [14] and viral suppression [15,16].In low-and middle-income countries there is limited information available for adolescents at a population level to inform national HIV strategic plans and programming [17].
In South Africa, routine data on ALHIV's health are also not readily available [3].Pertinent information on ALHIV can be sourced from a series of national population-based household surveys conducted since 2002 [2,[18][19][20].Consequently, we undertook a formative study which was part of a commissioned project entitled "Being ALHIV:' What do we know about Adolescents Living with HIV in South Africa?In this paper, we report key findings from quantitative analyses [21], focusing on HIV prevalence, incidence, and progress towards the UNAIDS 90-90-90 goals and HIV drug resistance (HIVDR).
The survey reports describe HIV testing assays for each SABSSM survey [2,[18][19][20].HIV prevalence was measured as the percentage of people tested and found to be living with HIV out of the target age group.HIV incidence was defined as the proportion of new infections acquired within the previous 12 months.Incidence data were only available for 2012 and 2017.The incidence estimation was based on a multi-assay algorithm incorporating the Limiting-Antigen Avidity assay (Maxim Biomedical, Rockville, USA), ART/ARV exposure and HIV viral load.Exposure to ARVs in the 2012 and 2017 surveys was measured by a qualitative determination of the presence of one or more ARVs in the testing panel, using a validated in-house method based on high performance liquid chromatography with tandem mass spectrometry [2,20].The testing panel accounted for ARVs used in first, second and third-line ART regimens in the country's public health HIV programme at the time of each survey [2,20].VLS was defined as a viral load of < 1000 copies of HIV RNA/ml.VLS is a measure of ART efficacy and is a proxy for ART adherence and HIV transmission risk [23].
Non VLS (VL ≥ 1000 copies/ml) samples were evaluated for HIVDR using Next Generation Sequencing [2,24,25].Amplification of a 1,084 base pair PCR fragment was performed as described [24].PCR products were sequenced on the Illumina MiSeq using MiSeq Reagent Kit v3 (Illumina Inc San Diego, CA, USA).Analyses of drug resistance mutations (DRMs) was based on the Stanford v8.0 algorithm, with a 10% prevalence detection threshold [26].We report on resistance detected overall and by drug class.
Awareness of HIV status was used to estimate.the90-90-90 UNAIDS' targets, which stipulates by 2020, 90% of PLHIV know their HIV status, 90% of those diagnosed with HIV receive ART, and 90% of all people receiving ART achieved VLS.The gaps in achieving the UNAIDS' targets were calculated based on the 90%-81%-73% cascade [27].

Statistical analyses
We used cross-sectional benchmarked weights per each survey wave and did not pool these for analyses.We calculated descriptive statistics for the sample characteristics by each indicator variable, using Pearson's Chi-square tests to detect differences among categorical variables, reporting 95% confidence intervals (CIs).Where CIs did not overlap, this was used to conclude statistical significance.Significance at p ≤ 0.05 are reported.All analyses except incidence estimates were performed in Stata version 15.0 (Stata Corp, College Station, Texas, USA), accounting for the complex survey design.The computational tools for the incidence estimates were developed by the South African Centre for Epidemiological Modelling and Analysis [28].

Discussion
There were approximately 360 600 ALHIV aged 10-19 years in South Africa in 2017.Female adolescents bear a disproportionate burden of HIV compared to their male counterparts.The gender differential is consistent with HIV incidence among adolescents living in sub-Saharan Africa [29] and can be attributed to sociocultural practices.These include early sexual debut, having multiple sexual partners, age disparate relationships, which co-exist with economic disparities and gender based power dynamics including gender-based violence [21,30].HIV incidence was largely unchanged in the 15-19 year age group between 2012 and 2017, possibly attributable to interventions targeting adolescent girls and young women.Although young people share concerns and needs as adolescents, HIV programmes must recognize their heterogeneity.Special attention must be given to female ALHIV needs, both in prevention and care programmes.
Progress towards the UNAIDS target of the first 90 at 62.3% being aware of their HIV status, is substantially lower compared to the general population, where 85% of Target 81% Target 90% Target 73% Target 90% Fig. 2 Gaps in achieving UNAIDS targets among adolescents living with HIV, South Africa 2017 based on the 90-81-73% cascade PLHIV aged 15-64 years knew their status [2].The reasons for younger adolescents not knowing their status are multi-faceted.For instance, they may have not been informed by a parent/caregiver, nor have a reason to test if they were not engaged in sexual activity, or if they are in good health.This can be partly attributed to complexities of parental disclosure due to fear and social stigma, perceived child's negative emotional reaction and concerns about the child being too young and immature to understand a diagnosis of being HIV positive [31].
We found low levels of ART use and low levels of VLS among ALHIV.Like other African region-based studies, ART uptake was lower for females than males [32,33].There might be several reasons for a gender differential in ART initiation, including female children being less likely to be taken for health care and later ART initiation in females, compared to males [32].Lastly, this study is among the first to estimate levels of HIVDR in ALHIV in a household survey in South Africa [2,25].Continued monitoring of HIVDR in ALHIV and in the general population is needed, particularly to detect pre-treatment HIVDR [34,35].

Conclusion
The high HIV prevalence among female adolescents and steady incidence highlight the need to prevent new infections among adolescents.There remains a need for improved access to quality sexual and reproductive health services that are youth-friendly and gender sensitive.

Limitations
This study is limited by the small sample size, particularly for those aged 10-14 years.Collection of blood specimens for determination of HIV status in this group are influenced by parents' or caregivers' reluctance to give consent as mentioned by other researchers in this area [36].The samples used to estimate incidence for this age group were small.Therefore, these estimates should be interpreted with caution.Nevertheless, this study contributes to the paucity of information available about ALHIV in South Africa.The 2017 wave is the most recent cross sectional, national HIV prevalence survey for South Africa, and the 6th wave is currently underway.