The only national dataset of HIV prevalence currently available in South Africa is the annual antenatal clinic (ANC) seroprevalence survey which is conducted in October each year. The anonymous, unlinked data is collected at approximately 400 of the public sector ANCs, distributed throughout the country. There are a number of inherent biases in this data which arise from measuring the status of pregnant women attending these facilities. These include:

- All clinic antendees are of child-bearing age

- All attendees are participating in unprotected sex (otherwise they wouldn’t be pregnant)

- Public ANC facilities serve 80% of all pregnant women in South Africa; 85.2% of these are black South African women - the sample therefore somewhat under-represents women from other race groups

- Public ANC clients tend to be from the lower socio-economic groups, and therefore the sample also under-represents women from middle to upper income brackets

- A further bias is that of sample size (some clinics test 500 women, others test 10)