Long-term usage patterns and clinical outcomes in a community-based differentiated antiretroviral therapy delivery programme in South Africa

Introduction: There is little data on long-term implementation and outcomes for people living with HIV (PLHIV) in differentiated antiretroviral therapy (ART) delivery programmes. We aimed to analyse usage patterns of and associated treatment outcomes in a community ART programme, within the Centralized Chronic Medicines Dispensing and Distribution programme, in South Africa over 3.5 years. Methods: We performed a retrospective cohort study among PLHIV on first-line ART who were eligible for community ART delivery between October 2016 and March 2019, from 56 urban clinics in KwaZulu-Natal, South Africa. Follow-up ended in March 2020. We measured referral rates and, among those referred, we characterized patterns of community ART usage using group-based trajectory modelling following referral. We used survival analysis to measure the association between community ART usage and loss-to-care (no visit for ≥365 days) and logistic regression to measure the association between community ART usage and viraemia (≥50 copies/ml). Results: Among the 80,801 patients eligible for community ART, the median age was 36 years, 69.8% were female and the median (interquartile range [IQR]) follow-up time was 22 (13–31) months. In total, 49,961 (61.8%) were referred after a median of 6 (IQR 2–13) months from first eligibility. After referral, time spent in community ART varied; 42% remained consistently in community ART, 15% returned to consistent clinic-based care and the remaining 43% oscillated between community ART and clinic-based care. Following referral, the incidence of loss-to-care was 3.93 (95% confidence interval [CI]: 3.71–4.15) per 100 person-years during periods of community ART usage compared to 5.75 (95% CI: 5.28–6.25) during clinic-based care. In multivariable models, community ART usage was associated with a 36% reduction in the hazards of loss-to-care (adjusted hazard ratio: 0.64 [95% CI: 0.57–0.72]). The proportion of patients who became viraemic after first community ART referral was 5.2% and a 10% increase in time in community ART was associated with a 3% reduction in odds of viraemia (adjusted odds ratio: 0.97 [95% CI: 0.95–0.99]). Conclusions: Community ART usage patterns vary considerably, while clinical outcomes were good. Promoting consistent community ART usage may reduce clinic burden and the likelihood of patients being lost to care, while sustaining viral suppression.