Educating religious leaders to create demand for medical male circumcision

Incident HIV infections refl ect the inadequacy of prevention strategies, and the failure to adequately use proven eff ective interventions. Following compelling evidence from three independent community trials in South Africa, Uganda, and Kenya, WHO and the Joint UN Programme on HIV/AIDS (UNAIDS) recommended that voluntary medical male circumcision (VMMC) should be part of a comprehensive HIV prevention strategy for heterosexually acquired infection among men. The eastern, central, and southern African countries have been prioritised for scaling up this intervention because of the high HIV burden and low prevalence of male circumcision in these countries. However, they have fallen short of their expected targets because of inadequate uptake of VMMC among men aged 15–49 years. Several research groups in sub-Saharan Africa are experimenting with diff erent approaches to substantially increase demand for VMMC. In The Lancet, Jennifer Downs and colleagues report a successful mixed-methods study in Tanzania that focused on educating religious leaders. They did a community cluster randomised trial in 16 villages, paired by proximity, and all received VMMC delivered by workers from the Tanzanian Ministry of Health services. Eight villages were randomly assigned to receive an education intervention for religious leaders in scientifi c, religious, and cultural aspects of male circumcision. The remaining villages (controls) did not receive the education intervention. Following completion of the randomised trial, focus group interviews were conducted with religious leaders in all villages. In the intervention villages, 52·8% (30 889 of 58 536) of men were circumcised compared with 29·5% (25 484 of 86 492) in the control villages (odds ratio 3·2 [95% CI 1·4–7·3]; p=0·006). In the intervention villages, 30·8% of men undertook VMMC because of church discussions, compared with 0·7% in control villages (p<0·0001). The interviews showed that religious leaders recognised the strong infl uence they had on their followers’ behaviour. The study had limitations, for example, it did not collect baseline data on circumcision status and religion in the study villages but relied on the country’s 2012 national census data, and it was assumed that the paired study villages had similar baseline characteristics. However, we are reassured that, despite the study limitations, the sensitivity analysis showed that the results were robust. Additional support for the positive outcome of a signifi cantly higher prevalence of circumcision in intervention than in control villages comes from the secondary qualitative fi ndings that religious leaders from the intervention villages felt empowered to teach their congregations about male circumcision after the educational seminars, and that many of the leaders described that more members of their congregations were seeking VMMC after church discussions of the topic. Additionally, wives and girlfriends in intervention villages were more likely to encourage their male partners to seek circumcision than those in the control villages (p=0·001). By contrast, people from the control villages had many negative opinions of circumcision, and misinformation and suspicion persisted in the control communities. The primary and secondary outcomes of this study suggest that social and structural determinants of behaviour can be infl uenced by incorporating religious leaders— including those opposed to the programme—as an integral part of community mobilisation in circumcision campaigns. Downs and colleagues estimated that their intervention strategy has the potential to encourage a million additional circumcisions, and potentially prevent 65 000–200 000 HIV infections in Tanzania, calling for worldwide attention to the innovative strategy used in this trial. Published Online February 14, 2017 http://dx.doi.org/10.1016/ S0140-6736(17)30318-5