The Soul City Institute for Health and Development Communication (SCIHDC) has been running Soul Buddyz Clubs (SBC) in primary schools in South Africa since 2003. SCIHDC wanted to determine the long-term impact of participation in SBC on previous members’ overall health, community participation and level of education attainment.
The clubs, whose members are between 8 - 14 years, focused on promoting positive attitudes and behaviours that related to HIV, such as self-efficacy for safer sexual behaviour, condom use, community activity, gender based violence and alcohol or drug use.
The sample comprised people who were SBC members between 2004 and 2008. We designed a retrospective cohort study design in order to determine the impact of being an SBC member. Cases comprised individuals aged 18-24 years who attended SBC's approximately 6-10 years ago (2004-2008). Controls were individuals aged 18-24 years who never attended SBC but grew up and attended in schools in the same communities. Our fieldwork partner, Social Surveys Africa, were able to interview 800 respondents (300 cases and 500 controls) between Gauteng and KwaZulu-Natal.
The challenges included having to find and recruit members retrospectively without a sample frame, establishing an appropriate control group, getting participants to come to a central venue for interviews and asking participants to get an HIV test.
Our approach was to first conduct a feasibility study to determine the likelihood of being able to locate previous SBC members and to determine the preferred control group. We used these findings to develop an innovative retrospective cohort study design to meet the study objectives.
We then used a multi-pronged recruitment approach to recruit the ex-SBC members. Methods included radio adverts, a Facebook group, WhatsApp and telephone calls to those identified through snowballing.
We incentivised referrals of previous SBCs to improve snowballing. We established where the majority of the ex-SBC members attended primary school and together with SSA we identified appropriate community controls using random sampling. SAA staff were at the fieldwork sites and ensured participants were given transport money and thank-you gifts. We ran competitions at the fieldwork sites to improve the response rate.
We were able to combine epidemiological expertise, in-depth sector knowledge and real world experience to solve a difficult evaluation problem for our client using an innovative study design. We found that ex-buddyz were more likely to have completed Grade 12, be involved in community activities and to have used condom at first sex. They were less likely to have sex before 15 years, to have had more than one partner in the past 12 months and have had multiple sexual partners in the past month.
Female ex-Buddyz were less likely to be HIV positive than controls. There was no effect on male ex-Buddyz, possibly because the sample size of HIV positive men in this age group was too small.
The results of the study will be used to determine whether school-based programmes have any lasting impact on HIV – an area where there is not much evidence. In addition, SCIHDC now has a list of previous members as well as an activate platform (Facebook group) through which it can communicate with them.