Published on June 30, 2009 by

SACEMA at the South African AIDS Conference

A new deal for South Africa

On 31 March 2009 John Hargrove, Director of SACEMA (with Brian Williams, Epidemiologist affiliated to SACEMA, as co-author) gave the plenary scientific talk at the opening of the 4th South African AIDS Conference in Durban with the title “The HIV Epidemic in South Africa – Provenance and Prospects” (1).

When one analyses the HIV prevalence rate in Africa, that is the proportion of people currently infected, it turns out that living in Southern (and to a lesser extent Eastern) Africa is associated with a large additional risk. This was found in a statistical model taking into account the proportion of males circumcised and the proportion of Muslims in any given country, both of which show a strong protective effect. It is suggested that this higher prevalence could be related to migration of men for employment, first to the mines and later to the cities, leaving their wives behind in the place of origin. This results in the destruction of the family unit, creating a breeding place for violent crime and sexually transmitted infections, including HIV/AIDS. A long-term solution would be changes in housing and employment practices, and rural development on various aspects including general health care and education. However, even in the scenario that these changes would occur, it would only be slow. Therefore medical interventions should be considered. But an effective vaccine has not yet been found, and although there are some promising results for an effective microbicide recently (2), it will take years before a (partially) effective one will be on the market. In the mean time two other options should be considered: First of all, a modelling study suggests that male circumcision reduces transmission, averaged over men and women, by about 40% (3). Bertran Auvert, who is one of the authors of the article on transmission probability of HIV and HSV-2 in this issue, wants to extend the randomised controlled trial of male circumcision to operational research by offering this intervention free of charge to all males in Orange Farm, South Africa. Another option would be annual HIV testing of the population and immediate antiretroviral therapy (ART) for everyone diagnosed with HIV, which could reduce new infections by 95% within ten years according to a modelling study (4,5). This could furthermore bring the incidence (new cases) of tuberculosis close to its pre-HIV level in ten years. The cost for HIV testing and early onset of ART is high but affordable and if the costs of people’s lives are included it is cost saving. John Hargrove therefore concluded by saying: “It will need strong and dedicated leadership but it could also be the beginning of A New Deal for South Africa”.

Estimating relative changes in HIV incidence: the BED method

John Hargrove also presented a case study of the application of the BED assay, an enzyme-based method that allows the time of HIV infection to be estimated for research purposes (6). In combination with mathematical modelling, this assay can be used for estimating HIV incidence from cross-sectional surveys. However, despite its popularity, there have been complaints that the method can be inaccurate. For example, the technique inappropriately identifies patients as recent infections when they have actually been infected for much longer than a year. The reason for this is not well understood. The ZVITAMBO data, a Zimbabwean dataset of 14,000 mothers who have been followed-up after giving birth, were used to look in detail to what extent this inaccuracy existed. It was demonstrated that unless one adjusts for the proportion of cases that are actually long-term cases, but identified as recent cases (false-recent cases) the BED method cannot reliably be used for estimating relative changes in HIV incidence. Another study presented at the conference by Thomas McWalter of the University of the Witwatersrand, indicates that the BED method has not been uniformly applied in publications (7). That results in different incidences rates and therefore standardized procedures are needed to make the rates comparable. Alex Welte, visiting research fellow at SACEMA, presented the latest progress in methodology for estimating incidence from cross sectional surveys by application of a test for recent infection at a special session hosted by UNAIDS (8).

These results have implications for public policy, because it is impossible to measure whether or not prevention and treatment interventions are having an effect, when no accurate estimate of the HIV incidence is available.

Mathematical models and HIV prevention

Farai Nyabadza, a post-doctoral fellow at SACEMA, presented a poster in which an introduction was given on the use of mathematical models of HIV transmission in the presence of interventions (media campaigns, condom use, HIV screening, and treatment) (9). When modelling HIV prevention strategies their impact can be quantified and predictions on disease outcomes can be made over a long period of time. Based on the outcomes of such a model, public health policies can be informed.

Wim Delva, PhD student at SACEMA, presented the development of a new type of mathematical model that assesses the impact of a combination of HIV prevention strategies in a population with complex and various sexual behaviour patterns (10). The model is designed to simulate the scaling up of condom use, antiretroviral treatment (preceded by voluntary counselling and testing), male circumcision and reduction in concurrency (having multiple relationships which overlap in time) over 10 years. The model takes into account that uptake and adherence to the different interventions varies over time, with age, sex, relationship status, duration of the relationship and use of other prevention methods.

There is evidence that Herpes Simplex Virus Type 2 (HSV-2) infection makes the chance of transmitting and getting HIV higher. Furthermore, two randomized controlled trials conducted in Africa showed that male circumcision reduces female-to-male sexual transmission of not only HIV, but also of HSV-2 (11,12). In the study presented in a poster, Guy Mahiane, visiting student at SACEMA, used data from the male circumcision trial from the Orange Farm (including 3274 men aged 18-24 years old who were followed-up for 21 months) (13). Guy developed a statistical model to assess the effect of prevention strategies (male circumcision and condom usage) and HSV-2 on HIV incidence. According to the study, HSV-2 makes the risk of getting HIV higher for males, and vice versa. Furthermore, these risks are shown to be reduced by male circumcision.

Controlling the epidemic by reducing intergenerational sex

Suppose HIV is introduced into a community. If partners always have the same age, or always the same age difference, HIV dies out in that community. However, HIV persists if there is variation in the age difference between partners. Carel Pretorius, PhD student at SACEMA, presented the development of an analytical model to find out how the age range of people’s sexual partners affects the extent of the epidemic (14). The results showed that the HIV epidemic will be sustained if, for example, 90% of people have sex with people of the same age while 10% of young people have sex with people who are ten or more years older than they are. Therefore, reducing the amount of intergenerational sex, especially in the case of young women having sex with older men, could contribute substantially to the control of HIV in South Africa.