Twelve-year mortality in adults initiating antiretroviral therapy in South Africa

dc.contributor.authorCornell, Mornaen_ZA
dc.contributor.authorJohnson, Leigh F.en_ZA
dc.contributor.authorWood, Robinen_ZA
dc.contributor.authorTanser, Franken_ZA
dc.contributor.authorFox, Matthew P.en_ZA
dc.contributor.authorProzesky, Hansen_ZA
dc.contributor.authorSchomaker, Michaelen_ZA
dc.contributor.authorEgger, Matthiasen_ZA
dc.contributor.authorDavies, Mary-Annen_ZA
dc.contributor.authorBoulle, Andrewen_ZA
dc.identifier.citationCornell, M., et al. 2017. Twelve-year mortality in adults initiating antiretroviral therapy in South Africa. Journal of the International AIDS Society, 20(1):21902, doi:10.7448/IAS.20.1.21902
dc.identifier.issn1758-2652 (online)
dc.descriptionCITATION: Cornell, M., et al. 2017. Twelve-year mortality in adults initiating antiretroviral therapy in South Africa. Journal of the International AIDS Society, 20(1):21902, doi:10.7448/IAS.20.1.21902.
dc.descriptionThe original publication is available at
dc.description.abstractIntroduction: South Africa has the largest number of individuals living with HIV and the largest antiretroviral therapy (ART) programme worldwide. In September 2016, ART eligibility was extended to all 7.1 million HIV-positive South Africans. To ensure that further expansion of services does not compromise quality of care, long-term outcomes must be monitored. Few studies have reported long-term mortality in resource-constrained settings, where mortality ascertainment is challenging. Combining site records with data linked to the national vital registration system, sites in the International Epidemiology Databases to Evaluate AIDS Southern Africa collaboration can identify >95% of deaths in patients with civil identification numbers (IDs). This study used linked data to explore long-term mortality and viral suppression among adults starting ART in South Africa. Methods: The study was a cohort analysis of routine data on adults with IDs starting ART 2004–2015 in five large ART cohorts. Mortality was estimated overall and by gender using the Kaplan-Meier estimator and Cox’s proportional hazards regression. Standardized mortality ratios (SMRs) were calculated by dividing observed numbers of deaths by numbers expected if patients had been HIV-negative. Viral suppression in patients with viral loads (VLs) in their last year of followup was the secondary outcome. Results: Among 72,812 adults followed for 350,376 person years (pyrs), the crude mortality rate was 3.08 (95% CI 3.02– 3.14)/100 pyrs. Patients were predominantly female (67%) and the percentage of men initiating ART did not increase. Cumulative mortality 12 years after ART initiation was 23.9% (33.4% male and 19.4% female). Mortality peaked in patients enrolling in 2007–2009 and was higher in men than women at all durations. Observed mortality rates were higher than HIVnegative mortality, decreasing with duration. By 48 months, observed mortality was close to that in the HIV-negative population, and SMRs were similar for all baseline CD4 strata. Three-quarters of patients had VLs in their last year, and 86% of these were virally suppressed. Conclusions: The South African ART programme has shown a remarkable ability to initiate and manage patients successfully over 12 years, despite rapid expansion. With further scale-up, testing and initiating men on ART must be a national priority.en_ZA
dc.format.extent10 pages
dc.publisherWiley Open Access
dc.subjectHighly active antiretroviral therapyen_ZA
dc.subjectHIV-positive persons -- Careen_ZA
dc.subjectHIV-positive personsen_ZA
dc.titleTwelve-year mortality in adults initiating antiretroviral therapy in South Africaen_ZA
dc.description.versionPublisher's version
dc.rights.holderAuthors retain copyright

Files in this item


This item appears in the following Collection(s)