Prevalence, associated factors and geospatial patterns of HIV drug resistance in South Africa among the virally unsuppressed adults: updated results from the 2017 national HIV household survey

Abstract

South Africa bears the highest HIV burden globally, with substantial antiretroviral therapy (ART) scale-up. HIV drug resistance (HIVDR) threatens progress toward UNAIDS 95–95-95 targets. We assessed national HIVDR prevalence and associated factors among virally unsuppressed adults in 2017, before dolutegravir (DTG)-rollout. Secondary analysis of the 2017 national HIV household survey, included 1 253 virally unsuppressed (defined as ≥ 1 000 copies/mL) adults (≥ 15 years) with HIVDR results. Survey-weighted analysis generated nationally representative estimates. Weighted multivariable logistic regression identified factors associated with HIVDR by treatment status. Province-stratified analyses examined sub-national disparities. Nationally, the study found high viral non-suppression estimated at 37.8% (95%CI: 35.2 to 40.2), and overall HIVDR prevalence of 26.3% (95%CI: 22.5 to 30.5; n = 369), driven predominantly by efavirenz-associated NNRTI resistance (14.0%). Prevalence was highest in Eastern Cape (36.6%) and rural farms (35.0%); provincial variation was not statistically significant. Stratified analysis revealed HIVDR concentration among divorced individuals (75.0%, 95% CI: 41.7–92.6) and those with high stigma (19.9%, 95% CI: 13.3–28.7) in Eastern Cape, whereas North West showed consistently low prevalence. HIVDR was higher among ARV-positive (43.3%) vs. ARV-negative individuals (21.1%, adjusted odds ratio [aOR] 4.44, 95% CI: 2.78–7.10, p < 0.001). In EFV-negative subgroup, the strongest risk factor was TB history (aOR = 4.94, 95% CI:1.46–16.67, p = 0.010), followed by secondary education (aOR = 2.24, 95% CI:1.15–4.35, p = 0.017) and stigma (aOR = 1.86, 95% CI:1.09–3.19, p = 0.023). Older age was associated with reduced odds of HIVDR (25–34 years: aOR = 0.46, 95% CI:0.23–0.93); ≥ 45 years: aOR = 0.44, 95% CI:0.19–0.99). For EFV-positive individuals, being never married increased HIVDR odds (aOR = 3.23, 95% CI:1.05–9.95, p = 0.042), while employment substantially reduced odds (aOR = 0.32, 95% CI:0.10–0.97, p = 0.044). Elevated HIVDR among youth (15–24 years) likely reflects perinatal infection, prolonged paediatric ART, and treatment interruptions during transition to adult care. The 2017 pre-DTG baseline reveals high HIVDR (26.3%) shaped by clinical (ART, TB) and socio-structural (unemployment, stigma, marital status) factors. Distinct provincial and age-specific risk profiles, including youth with paediatric treatment histories, underscore the need for geographically tailored interventions, expanded resistance testing and youth-focused adherence and transition programmes.