36 Gordon Square
London WC1H 0PD

T +44 (0) 20 7958 8251
admin@lidc.bloomsbury.ac.uk

Combining Epidemiology and Social Science to Understand HIV Transmission

Recognising the role of men in positions of power could improve the traditional sex worker-centred analysis of the spread of HIV, according to a new study. Professor Charlotte Watts, of the London School of Hygiene and Tropical Medicine (LSHTM), delivered the exciting analysis during a presentation at LIDC. Watts, Director of the Gender Violence and Health Centre at LSHTM, highlighted social science insights that demonstrate the transitory nature of sex work for some sex workers. These mobility patterns are likely to have significant implications that may alter the current understanding of the transmission of HIV and the approaches to prevention. She showed, for example, how policemen and pimps may, in some circumstances, have more sexual partners over a five-year period than sex workers who may only engage in sex work for one year during the same period.

Watts said: “What is exciting about this work is that it starts to give us a different picture. The temporality of behaviour has been missed in epidemiological concepts, but social science reveals this complexity.”

The talk, entitled “Reconstructing the Core Group: Using Social Science Findings to Challenge Epidemiological Theories of HIV Transmission”, was based on the work of epidemiologists, mathematicians and social scientists from LSHTM (Watts, Cathy Zimmerman, Anna Foss, Cathy Lowndes, Mazeda Hossain and Peter Vickerman). The seminar was the first in a series called the” Social Science of Infectious Disease”, run jointly by LIDC and the London School of Economics.

Epidemiological explanation
Watts outlined the enormity of the HIV epidemic, which has claimed 25 million lives in the last 25 years, and spoke about the traditional epidemiological analysis used to explain it.
In epidemiology, the basic reproductive rate - R0 - gives a measure of the rate at which infection will spread within a susceptible population. Mathematically, it is expressed as R0 = (ßcD). (ß is the probability of infection transmission per partnership, c is the  rate of sexual partner change and D is the average duration of HIV infectivity). Therefore, if R0 is greater than one, the infection will spread, and if it is less than one it will die out. The ‘core group’ is conceptualised as being the individuals in a population who will infect more than one person over the duration of their HIV infection, and this ‘core group’ has been recognised as including sex workers and injecting drug users. This, in turn, has led to a strong research focus on commercial sex around the world. However, Watts demonstrated there are limitations to this rather simplistic ‘core group’ approach because it overlooks underlying contextual  factors, such as the dynamics of sex work settings, and it stigmatises female sex workers by emphasising their nearly singular role in the spread of HIV.

Social science insights
Unlike epidemiology, social science stresses the complexity and heterogeneity of women’s engagement in commercial sex, which takes place in varied settings, frequently or occasionally, with a range of men (policemen, boyfriends and pimps) and under varying levels of coercion. Watts highlighted the mobility and temporary nature of some sex work, for example, those who are no longer deemed profitable by pimps (e.g., as they get older), such as  in Vietnam, due to the high demand for young girls. She also spoke of how pimps may rape sex workers to initiate them into sex work, and presented data showing the ‘phenomenal level’ of sexual violence in Cambodia.  These concerns revealed the importance of understanding power relations between men and women, especially in the sex work setting. Watts added: “We need to recognise the power structures underlying sex work”.

New model and more research questions
The significance of the temporality of sex work was tellingly shown when Watts compared two scenarios. The first example, grounded in epidemiology, was based on a sex worker having 15 new clients per month and having sex with a pimp and a policeman over a five-year period. The total number of sexual partners in such a setting would be 902, far more than the 240 partners a policeman would have over the same period if he had sex with four new sex workers per month. However, Watts proceeded to reveal the very different situation if transitory considerations based on social science were included in the equation. For example, if the same sex worker only remained in that situation for one year within that five-year period she would have 182 (15x12) sexual partners, less than the policeman’s figure of 240. Watts drew on these numbers to pose a crucial question: “Could men who remain in the sex work setting be a neglected sustaining population?”

She said men in positions of power, such as pimps and policemen, may play a central role in infecting new sex workers , although they have not really been studied in behavioural research. Therefore, new research questions focusing on this male group arise from the supposition drawn from the two scenarios. During the presentation on 29 October 2008, Watts said it may be better to think of D in the R0 = (ßcD) equation as the duration in a particular risky setting rather than the average duration of HIV infectivity. She added: “A static picture is inherent in a lot of the modelling and we rarely collect data on or publish how long women have been in sex work. We have been trying to talk across disciplines and are researching questions which we would not be researching if we had kept in our silos”.

By Guy Collender, Communications Officer, LIDC