The first section is based on data from the Department of Health (2003)' National HIV and Syphilis Sero-prevalence Survey of Women Attending Public Antenatal Clinics in South Africa - 2002, Summary Report' This report is the 13th in a series of studies, and looks at data predominantly taken from antenatal clinics and examines the impact of HIV amongst the sexually active population of South Africa. It looks at information taken from 396 sampling sites, and samples from 16,587 women.

The second section is based on information taken from the Nelson Mandela/HSRC Study of HIV/AIDS, South African National HIV Prevalence, Behavioural Risks and Mass Media Household Survey 2002. The study contacted 13518 people, of whom 9963 people agreed to be interviewed and 8840 people agreed to give samples of saliva, which were taken for testing.

The two studies produce different results for estimated HIV prevalence because they look at different populations.

Limitations of the two studies

The Department of Health study takes data from antenatal clinics, and as a result of this is able to draw conclusions about HIV prevalence amongst sexually active women, but these conclusions cannot be applied to other groups in the population, such as non-sexually active women, the elderly or children. The estimates, however, do provide a good indication of prevalence rates amongst sexually active members of the population.

The Nelson Mandela study has tried to examine a representative cross-section of South African society but uses a relatively small sample. This study can, however, give an idea of HIV prevalence levels amongst such groups as children and the elderly. 3555 people of those originally contacted by the survey declined to be interviewed and 4678 declined to give a sample for testing. One reason for this could be that in South Africa, people are generally reluctant to take part in surveys, partly due to fear of crime. Another reason, however, might have been a reluctance to take a HIV test, perhaps because the subjects knew themselves to be at risk of a positive result.

The department of Health study shows a higher prevalence rate because it looks at the sexually active population. This does not mean that the rate will be as high amongst younger or older sections of society, or amongst men. The Nelson Mandela study looks at the whole population, but because it includes young and old people (who are less at risk of infection), the overall prevalence rate is consequentially lower. Seen together, these two studies provide a clearer picture than either of them viewed alone.

In 2001 it was estimated that in South Africa, 4.7 million people are living with HIV, of whom 189,000 are babies. Based on extrapolation of the results of this 2002 antenatal survey, the Department of Health estimated that 5.3 million South Africans were HIV positive by the end of 2002. It is estimated that 91,271 babies were infected with HIV during 2002 (250 a day) by mother-to-child transmission, which, with provision of the correct drugs, is almost completely preventable. These high prevalence rates show what a significant problem HIV/AIDS is in South Africa with enormous social economic and development implications. The rates also indicate the future burden of HIV associated disease and the difficulties faced by the health system in coping with the provision of adequate care and support.

These estimates have been made using a model developed by the South African Department of Health for this purpose, making sure that at antenatal sites across the country, new attendees have an equal chance of being tested. The surveys have shown that until 1998 South Africa had one of the fastest growing epidemics in the world. The 2002 survey confirms the trend seen since 1998 that this is no longer the case, and whilst the magnitude of the HIV/AIDS epidemic should not be underestimated, it does seem that the rapid growth of the South African epidemic may be slowing down. The stabilization of the epidemic is supportive by the prevalence rates in women/girls under 20.

The tables below show the estimated HIV prevalence by Province and by Age Group. The HIV prevalence is the percentage of people tested in each group who were found to be infected with HIV, The tables only display the average percentage of the minimum and maximum rates for HIV prevalence. For the confidence interval, see the original study.

In 2002, the province that recorded the highest HIV rate amongst antenatal attendees is KwaZulu-Natal, which had a rate of 36.5%, an increase of 3% since 2001. The next highest HIV levels were found in Gauteng (31.6%) and Free State (28.8%). The increases shown in KwaKulu-Natal, Gauteng, North West, Eastern Cape and Limpopo and the decreases shown in Free State, Mpumalanga and Northern Cape were not said to be statistically significant as they could be due to sampling variations. If patterns are emerging they will become apparent as they are monitored over the coming years.

An estimated 34.5% of women aged from 25 - 29 are infected with HIV, making this the age group with the highest prevalence rate. 29.5% of women aged 30 - 34 are estimated to be infected, and 29.1% of women aged 20 - 24 years are estimated to be infected. The rest of the age groups have lower prevalence rates. There were increases in the HIV prevalence in the 25 - 29 and 30 - 34 years age-groups which were statistically significant, and the increase in the 40+ years age group was statistically very significant. The prevalence rates in the <20 years age group suggests a continued stabilisation, which is encouraging.

About the survey and the survey model

The survey was conducted across all nine provinces from the 1st to 31st October 2002 with the aim of making nationally representative estimates of HIV prevalence in South Africa. The study was an anonymous, unlinked cross-sectional survey of pregnant women who attend antenatal clinics in the public health sector of South Africa. All pregnant women who attended antenatal care for the first time during their current pregnancy were eligible for inclusion in the survey.

16,587 women participated in the survey at 396 sampling sites, compared to 16,743 pregnant women and 421 sites in 2001. Based on the 16,587 blood samples that were tested, it is estimated that nationally 26.5% of pregnant women were infected with HIV by the end of 2002. This compares with a prevalence rate of 24.5% in 2000 and 24.8% in 2001

An antenatal survey is an internationally recommended surveillance tool to estimate HIV in populations. Whilst it has inherent limitations it is still considered the most useful tool to assess HIV prevalence in areas of high HIV prevalence. Pregnant women are sexually active, constitute an easily identifiable, accessible and stable population, and are more likely than other groups to be representative of the general population. In addition, pregnant women attend health care facilities where blood is drawn as part of routine medical services.

The antenatal survey gives estimates of HIV prevalence among pregnant women in South Africa but no direct information is available on infection rates among non-pregnant women, men, newborn babies and children. In order to obtain an estimate of the approximate number of South Africans infected with HIV, projections are therefore made to extrapolate the total number of people infected. These national estimates need to be interpreted with caution as there are inherent limitations in the methodology used..

Another study, the Nelson Mandela/HSRC study of HIV/AIDS (2002), has suggested that the antenatal survey has limitations, as it looks only at sexually active women. The Nelson Mandela study suggests that a survey which looks only at prevalence of HIV/AIDS amongst sexually active women will have difficulty drawing conclusions about prevalence amongst other sections of the population.

The Nelson Mandela survey is intended to be a 'household' study, looking at a proportional cross-section of society. A total of 14450 people were selected, of whom 4001 were children, 3720 youths and 6729 adults. Of these, 93.6% responded, and 65.4% of those who responded agreed to give a specimen for a HIV test.

The figures below show the HIV prevalence estimates produced by the Nelson Mandela study.