Modelling HIV incidence and survival from age-specific seroprevalence after antiretroviral treatment scale-up in rural South Africa

OBJECTIVE: Our study uses sex-specific and age-specific HIV prevalence data from an ongoing population-based demographic and HIV survey to infer HIV incidence and survival in rural KwaZulu-Natal between 2003 and 2011, a period when antiretroviral treatment (ART) was rolled out on a large scale. DESIGN: Catalytic mathematical model for estimating HIV incidence and differential survival in HIV-infected persons on multiple rounds of HIV seroprevalence. METHODS: We evaluate trends of HIV incidence and survival by estimating parameters separately for women and men aged 15–49 years during three calendar periods (2003–2005, 2006–2008, 2009–2011) reflecting increasing ART coverage. We compare model-based estimates of HIV incidence with observed cohort-based estimates from the longitudinal HIV surveillance. RESULTS: Median survival after HIV infection increased significantly between 2003–2005 and 2009–2011 from 10.0 [95% confidence interval (CI) 8.8–11.2] to 14.2 (95% CI 12.6–15.8) years in women (P < 0.001) and from 10.0 (95% CI 9.2–10.8) to 14.0 (95% CI 10.6–17.4) years in men (P = 0.02). Our model suggests no statistically significant reduction of HIV incidence in the age-group 15–49 years in 2009–2011 compared with 2003–2005. Age-specific and sex-specific model-based HIV incidence estimates were in good agreement with observed cohort-based estimates from the ongoing HIV surveillance. CONCLUSION: Our catalytic modelling approach using cross-sectional age-specific HIV prevalence data could be useful to monitor trends of HIV incidence and survival in other African settings with a high ART coverage.