Trajectories and predictors of HIV care retention among individuals receiving ART in rural South Africa: a group-based trajectory modelling analysis.

Background: Successful retention in care of people living with HIV remains a challenge and a cornerstone of ending the epidemic. A better understanding of retention predictors could guide an evidence-based approach to target interventions. We sought to characterise HIV care retention trajectories among individuals receiving antiretroviral therapy (ART) in a rural South Africa setting, and to determine factors associated with those trajectories. Methods: We conducted a population-based cohort study of individuals receiving ART in ten health care facilities within the Agincourt Health and Socio-Demographic Surveillance System site in Mpumalanga, South Africa, in 2015-18. We used group-based trajectory modelling to identify clusters of individuals with similar retention trajectories and assessed the association between socio-demographic factors and trajectory groups using multinomial logistic regression. Results: Among 1689 individuals receiving ART during the study period, five distinct trajectory groups were identified: 30.8% had gradually decreasing retention over time, 10.2% had late increasing retention, 20.7% had early increasing retention, 7.8% had early decreasing retention, and 30.5% had consistently high retention. Individuals in the consistently high retention group were more likely to be female and aged ≥40 years. In contrast, those in the early decreasing retention group were more likely to be male, aged <30 years, and with temporary resident status. Individuals in the early increasing retention group were more likely to be from villages included in a HIV Treatment as Prevention community mobilisation study. Education, marital status, and socioeconomic status were not significantly associated with group membership. Months on ART were weakly associated with group membership. Conclusions: Five distinct retention trajectories were observed and associated with specific sociodemographic factors. Our study offers a data-driven approach to inform the design of targeted interventions to improve HIV care retention. Interventions and policies addressing socioeconomic and system-level factors are essential to achieving better outcomes in high-burden areas.