Long-term adherence to antiretroviral therapy in a South African adult patient cohort: a retrospective study

BACKGROUND: South Africa has the highest HIV prevalence and supports the largest antiretroviral therapy (ART) programme globally. With the introduction of a test and treat policy, ensuring long term optimal adherence to ART (≥95%) is essential for successful patient and public health outcomes. The aim of this study was to assess long-term ART adherence to inform best practices for chronic HIV care. METHOD: Long-term ART adherence was retrospectively analysed over a median duration of 5 years (interquartile range [IQR]: 5.3–6.5) in patients initially enrolled in a randomised controlled trial assessing tuberculosis and HIV treatment integration and subsequently followed post-trial in an observational cohort study in Durban, South Africa. The association between baseline patient characteristics and adherence over time was estimated using generalized estimating equations (GEE). Adherence was assessed using pharmacy pill counts conducted at each study visit and compared to 6 monthly viral load measurements. A Kaplan Meier survival analysis was used to estimate time to treatment failure. The McNemar test (with exact p-values) was used to determine the effect of pill burden and concurrent ART and tuberculosis treatment on adherence. RESULTS: Of the 270 patients included in the analysis; 54.8% were female, median age was 34 years (IQR:29–40) and median time on ART was 70 months (IQR = 64–78). Mean adherence was ≥95% for each year on ART. Stable patients provided with an extended 3-month ART supply maintained adherence > 99%. At study end, 96 and 94% of patients were optimally adherent and virologically suppressed, respectively. Time since ART initiation, female gender and primary breadwinner status were significantly associated with ≥95% adherence to ART. The cumulative probability of treatment failure was 10.7% at 5 years after ART initiation. Concurrent ART and tuberculosis treatment, or switching to a second line ART regimen with higher pill burden, did not impair ART adherence. CONCLUSION: Optimal long-term adherence with successful treatment outcomes are possible within a structured ART programme with close adherence monitoring. This adherence support approach is relevant to a resource limited setting adopting a test and treat strategy.