4338Impacts of intimate partner violence and sexual violence on antiretroviral adherence among adolescents living with HIV in South Africa

L. Cluver * (1,2), S. Zhou (3,4), M. Orkin (5), S. Dzumbunu (3), F. Meinck (6,7), N. Langwenya (1,3), M. Vicari (8), L. Sherr (9), E. Toska (3,10) (1) University of Oxford, Department of Social Policy and Intervention, Oxford, United Kingdom, (2) University of Cape Town, Department of Psychiatry and Mental Health and Centre for Social Science Research, Cape Town, South Africa, (3) University of Cape Town, Centre for Social Science Research, Cape Town, South Africa, (4) University of Cape Town, Department of Public Health and Family Medicine, Faculty of Health Sciences, Cape Town, South Africa, (5) University of the Witwatersrand, Medical Research Council Development Pathways to Health Research Unit, School of Clinical Medicine, Johannesburg, South Africa, (6) University of Edinburgh, School of Social and Political Science, Scotland, United Kingdom, (7) University of the Witwatersrand, School of Public Health, Johannesburg, South Africa, (8) International AIDS Society, Geneva, Switzerland, (9) University College London, Institute for Global Health, London, United Kingdom, (10) University of Oxford, Social Policy and Intervention, Oxford, United Kingdom

BACKGROUND: We are failing to reach 95-95-95 for adolescents living with HIV (ALHIV). Adolescents in Sub-Saharan Africa are exposed to high rates of sexual violence and intimate partner violence (IPV). However, evidence on associations of sexual violence and ART adherence remains limited, with only three cross-sectional studies globally.
METHODS: We conducted a longitudinal cohort, with interviews and clinical records from 1046 ALHIV aged 10-19 years, recruited from 53 government health facilities in South Africa's Eastern Cape (2014-2018; 57% female, 90% uptake, 94-97% retention, 3.4% mortality).Ethical approvals were given by the University of Cape Town, University of Oxford, provincial government and health facilities. We used a repeated-measures random effects model and marginal predicted probabilities to assess multivariable associations of self-reported sexual violence and IPV with ART adherence, validated against viral load suppression (<50 copies/ml). We fitted moderation models by gender.
RESULTS: 51% of adolescents reported consistent ART adherence. ART adherence was associated with viral suppression (aOR 1.49, CI:1.03-2.14, p=0.033) controlling for age, sex, location, poverty, housing, and HIV acquisition mode. Exposure to IPV was associated with lower ART-adherence (aOR 0.39, CI:0.21-0.72, p=0.003), and so was sexual violence (aOR 0.54, CI:0.29-0.99, p=0.048). Marginal predicted probabilities showed that adolescents with no sexual violence or IPV exposure had a 72% (CI:0.70-0.74) probability of ART adherence compared to 38% (CI:0.20-0.56) for those exposed to both sexual violence and IPV. Moderation showed similar impacts of violence by gender.
CONCLUSIONS: Effective sexual violence prevention and post-violence care are essential in supporting adolescent ART adherence. There is now increasing evidence of effective services across sectors, with systematic reviews identifying parenting programmes, classroom and community-based programmes in sub-Saharan Africa, and social protection such as government cash transfers in reducing sexual violence. There is an urgent need to link violence prevention and adolescent HIV services.

Figure 1: Adjusted predicted probabilities of past-week adherence
'