Estimates and predictors of HIV viral non‐suppression in South African adults on antiretroviral treatment

INTRODUCTION: Viral suppression estimates are essential for monitoring the performance of HIV programmes. South Africa introduced dolutegravir (DTG)-based antiretroviral therapy (ART) in 2019. We sought to generate updated estimates of viral suppression in South African adults on ART and investigate predictors of viral non-suppression. METHODS: This retrospective cohort study used data from seven South African cohorts participating in the International epidemiology Databases to Evaluate AIDS collaboration. Three main analyses were performed using a viral suppression threshold of 1000 HIV RNA copies/ml. In the first analysis, we fitted a logistic regression model using the full data from the study period (2005-2023). Then, in two causal analyses, we used logistic regression with inverse probability weighting to assess the effects of starting ART on DTG-based regimens (as opposed to starting on non-DTG-based ART) and switching to DTG while virally suppressed (compared to remaining on non-DTG-based ART). In sensitivity analyses, we reduced the suppression threshold to 400 copies/ml and excluded those with missing baseline CD4+ cell count measurements. RESULTS: There were 380,720 participants contributing 2,090,912 person-years of observation. Most participants were female (64.7%), and the median age at ART initiation was 35.0 years (interquartile range 28.9-42.3). Viral suppression increased over time, reaching 95.9% in 2023. Twenty-one percent of participants either started ART on DTG-based regimens (7.1%) or switched to DTG-based regimens from a virally suppressed state (14.0%). DTG-based ART was protective against viral non-suppression in both causal models, with adjusted odds ratios of 0.54 (95% confidence interval [CI] 0.48-0.61) and 0.36 (95% CI 0.32-0.39) for those initiating ART on DTG and those switching to DTG, respectively. A history of ART interruption was strongly associated with viral non-suppression, with adjusted odds ratios ranging from 2.49 to 4.55. The odds of non-suppression decreased with increasing age, increasing duration on ART and increasing baseline CD4+ cell count. Results were consistent across sensitivity analyses. CONCLUSIONS: DTG-based regimens improve viral suppression among both ART-naïve individuals and those transitioning while suppressed. ART interruptions pose a risk to the sustained success of ART programmes and may further impede efforts to recover from the impacts of recent funding cuts.