The association between HIV treatment interruptions and viral suppression during the early treatment period: retrospective cohort study in South Africa

Résumé fourni par la source

ABSTRACT Background The first six months after antiretroviral therapy (ART) initiation for HIV is a high-risk period for treatment interruptions that may compromise viral suppression (VS). Recent research in South Africa suggests that >40% of patients interrupt care for >28 days during the early treatment period. The quantitative association between early treatment interruptions and VS at 6 and 12 months remains unclear. Methods We enrolled adults (≥18 years) initiating ART from 1 January 2018 to 7 November 2024 with at least 14 months’ followup in South Africa’s national ART database (TIER.Net) from 24 public sector facilities in four provinces. Engagement in care during months 0-6 and 7-12 was classified as continuous (no interruptions >28 days), cyclical (at least one interruption >28 days but returned to care within follow up period), or disengaged (>28 days late without return), based on completed and scheduled visit dates. Modified Poisson regression was used to estimate adjusted risk ratios (aRRs) for VS (<50 copies/mL), adjusting for age, sex, initiation year, regimen, engagement pattern, and baseline CD4 count. Findings Among 57,553 participants (66% female; median age 33 years), 49% and 42% were continuously engaged at 6 and 12 months, respectively; 22% and 17% were cyclically engaged at the same time points. 54% of continuously engaged participants achieved 6-month VS compared with 34% of those with cyclical engagement (aRR 1.60 95% CI 1.55-1.64). At 12 months, 56% of continuously engaged individuals and 40% of those cyclically engaged were suppressed (aRR 1.38 95% CI 1.34-1.42). VS was also associated with dolutegravir-based regimens, later ART initiation year, baseline CD4 count ≥200 cells/uL, female sex, and older age. Interpretation Even relatively brief treatment interruptions during the first year of ART were associated with substantially lower viral suppression. Preventing early interruptions should remain a programmatic priority to improve treatment outcomes. Funding Funding for this study was provided by the Gates Foundation under INV-031690 to Boston University.