Background:
Differentiated antiretroviral therapy (ART) delivery programmes are being rolled out globally but there is little data on long term implementation and outcomes for people living with HIV (PLHIV). We describe patterns of exposure to community ART delivery and associated treatment outcomes in the South African Centralized Chronic Medicines Dispensing & Distribution programme over 3.5 years.
Methods:
We performed a retrospective cohort study among PLHIV on first-line ART who were referred for community ART delivery between Oct 2016 – Mar 2019 from 56 clinics in KwaZulu-Natal, South Africa. Follow up ended in Mar 2020. We used group-based trajectory modelling to characterize patterns of exposure to community ART following referral, and survival analyses to measure the association between time-varying exposure to community ART and time to loss-to-care, defined as no clinic visit for >365 days. We used logistic regression with generalized estimating equations to quantify the association between proportional exposure to community ART 12 months prior to viral load measurement and viraemia ( >50 cps/mL) among those in care. Models of loss-to-care and viraemia adjusted for baseline patient age, gender and time on ART.
Results:
Among the 37,596 patients referred to community ART, 68.6% were female with a median (interquartile range (IQR)) age of 38 (32-45) years, and 99.3% were on tenofovir-based regimens. Median (IQR) follow-up time was 656 (430-841) days. Time spent in community ART varied during follow-up; ~40% remained consistently in community ART following referral, ~ 20% returned to clinic-based care after their first referral visit while the remaining 40% oscillated between community ART and clinic-based care. The incidence of loss-to-care was 4.46 per 100 person-years and community ART exposure was associated with a 41.1% (95% confidence interval (CI): 34.1%-47.3%) reduction in the hazard of loss-to-care after adjusting for other covariates. Among those with at least one viral load measured after referral (Nf33,378), 5.4% became viraemic. A 10% increase in the proportion of time spent in community ART in the 12-month period before viral load measurement was associated with a 4% (95% CI: 2%-5%) reduction in the adjusted odds of viraemia.
Conclusions:
Community ART exposure patterns vary considerably after referral into the programme. Promoting consistent programme adherence will reduce clinic burden and the likelihood of patients being lost to care, whilst sustaining viral suppression.