Date Published: June 8, 2010
Publisher: Public Library of Science
Author(s): Rachel Jewkes
Abstract: Rachel Jewkes discusses disappointing results from a school-based sexual health intervention study in Tanzania and their implications for future health education programs.
Partial Text: Sub-Saharan Africa remains the global hub of the HIV epidemic. Seventy percent of new HIV infections occur in the region. And because the sexually transmitted epidemic starts among women and men in their teenage years , promoting the sexual health of adolescents is a substantial health priority. In light of this, the findings of the MEMA kwa Vijana trial in Tanzania  and the long-term follow up of participants in its school-based intervention, published this week in PLoS Medicine, are very disappointing. The key questions we have to ask now are: In light of current knowledge about behaviour change, are these findings surprising? And, what are the implications for the next generation of sexual health interventions?
In their paper reporting the impact of the intervention 9 years after it was implemented, David Ross et al.  note that this intervention, although remarkable for the extent and rigor of its evaluation, now joins the ranks of a much larger set of school-based interventions in sub-Saharan Africa that have been shown to be ineffective in changing sexually risky behaviour. Plummer et al.  present a stark picture of the school environment that clearly shows how critically important context is for the success of HIV prevention interventions. This is echoed by a growing body of educational theory that seeks to explain why many HIV interventions in schools have been ineffective, and it points to the importance of transforming schools, and educational policy, as a whole. An important contribution here has been made by Unterhalter  and elaborated in Morrell et al. . After a decade of empirical research on gender and HIV in schools in South Africa, they reflect that policy approaches to gender, violence, and HIV have three dimensions, which at best hold the potential for being profoundly empowering for learners. These are: interventions, which focus, for example, on introducing ideas, giving information, providing skills, or punishing transgressive behaviours; institutions, with their policies able to shape social relations and ways of working; interactions, which involve all social relationships in schools, and provide the daily space in which gender identities and relations are enacted and where values are conveyed and may be contested. Unterhalter and Morrell et al. argue that empowerment requires good practice in all three of these dimensions, yet there has been an overwhelming focus on some interventions, more limited attention to institutions through laws and policies and their implementation, and a substantial neglect of attention to interactions. This narrows the frame of many interventions, and may explain the apparent ineffectiveness.
The findings of Ross et al.  should not be interpreted as showing that school-based health education programmes in sub-Saharan Africa do not work, but rather that we must learn to do them better. We do not know yet about how to deliver effective HIV prevention through schools. This study reminds us that we cannot safely assume that evidence of attitude change and some behaviour change found in evaluations will translate into long-term protection against HIV or other sexually transmitted infections. If we are to develop an evidence base for programming in schools, substantial resources are needed for evaluations of promising programmes against robust outcomes. We need more research, more large studies, and much more resources.