New HIV infections (per 1000 uninfected population) | South Africa – yearly data, chart and table

Spectrum modelling is used for the data presented here which incorporates programme data, surveillance data, survey data and region-specific assumptions about the HIV epidemic. Longitudinal data on individuals newly infected with HIV would be the most accurate source of data to measure HIV incidence, however these data are rarely available for representative populations. Special diagnostic tests in surveys or from health facilities can also be used to obtain data on HIV incidence but these require very large samples to accurately estimate HIV incidence and the latter are also rarely representative. Alternative methods of measures include household or key population surveys with HIV incidence-testing, or routine surveillance among key populations. Data sources are compiled all year long. The spectrum models are created in the first three months of every year and finalized by May. The estimates are produced by a team of national experts consisting of ministry of health, national AIDS advisory groups and development partners. The results are signed off on by senior managers at the ministries of health. Data compilers After the data review process, the national experts share their results with UNAIDS who compiles the data for all countries and calculates regional and global estimates.