The session ‘Social Sciences and Intervention – putting theories into practice’ provide a series of examples and an opportunity to debate some key issues on the utility and methodologies of social science as applied to HIV prevention treatment and care.  The session started with an overview of structural intervention presented by Carr (TUAD0101), one of the co-chairs. He emphasized quite effectively that it is important to use social sciences to understand social structures such as gender, stigma and social exclusion, as it informs us about the drivers of the responses to HIV and why programs work or not . He suggested that interventions that take structural issues into account by using the ecological approach enhances our ability to understand and influence individuals’ behaviors within the household, community and cultural contexts. He commented that social sciences are important but social sciences evidences are often seen as being inferior to clinical data. The presentation however, did not define clearly what structural interventions are. One question from the floor asked for clarification and the answers may still remain vague.

Kerrigan (TUAD0102), the second speaker presented  her multi-level intervention targeting FSW in Brazil. At the individual level, the program tried to provide the clients with clinical and psychological services. At the community level, it organized collective activities, peer educators, workshops, cultural performance and workshops .  These were carried out in collaboration with governmental and community organizations. Issues such as equality and stigma were considered. The 420 participants were followed up for 15 months. It was concluded that improvements were observed in social cohesion and social network indicators, as well as on prevalence of condom use with some types of sexual partners. The speaker emphasized the importance of changes to  the social environment in order to improve HIV-related responses.

Maikweki (TUAD0103) then introduced their series of behavioural change communication (BCC)  programs targeting young people in Kenya, which evolved since 2001. The evaluation performed in 2005 found that condom use increased slightly (54 to 58%), but definition of condom use was not very clear).  Risk perceptions remained very low. It was concluded that knowledge would not induce behavioral changes. A fact that has been well established in the social science literature for over 20 years.  The program was revised in 2005, emphasized more on changing risk perceptions, self-efficacy, social norms. It was evaluated in 2009, finding that condom use rates increased from 54 to 69%. The study design and presentation of results were however, not very clear, and was challenged  by a member of the audience from the floor.

Jemmott (TUAD0104) presented a typical RCT on an intervention aiming at promoting condom use among 1070 HIV discordant African American couples in 4 cities. Workshops were run based on social cognitive theory and ecological theory, attempting to change norms, self-efficacy and the like. The 12-month follow up data has a high retention rate (almost 90%), reporting improvements in condom use rates but not STD incidence. Monetary incentives were given to the participants.

 Reddy's project covered over 40000 children affected by HIV in India. It provided health education, nutrition and psychosocial support and a safety net. The details of the operation were introduced. Evaluation showed that the testing rates increased from 19-58% and loss to follow up for ART services was below 2%.

There are some interesting observations. First, it is encouraging to see that all speakers emphasize  the importance of social sciences and the message is clear and consistent. However, disturbingly an audience member  who is an HIV clinician openly questioned the usefulness of social sciences for HIV prevention. It seems that we still have to convince many HIV workers that an inter-disciplinary approach is required. Second, all speakers mentioned  structural intervention and endorsed this approach, but it seems that they may mean different things and a stronger clearer conceptual framework is required. Third, it is encouraging to see that we are now more used to including cognitive behavioral determinants such as self-efficacy as part of the interventions. Finally, there were debates in the meeting about what comprises evidence. The presentation on BCC was criticized by the lack of stringent study design. A critical debate on the tools, methodology and utlity of social science was informative - there seems to be a wide view of what constitutes evidence from a social science perspective.  Extreme views on both sides were expressed, but it seems that good quality evidence, from competent qualitative and quantitative studies feed the knowledge base. Consensus building is required before social sciences can serve HIV work effectively.