RESUMEN
BACKGROUND: Bangladesh reports one of the highest rates of intimate partner violence (IPV) in the world. Despite wide recognition of IPV as an important public health and human rights issue, evidence for IPV prevention is still inadequate. Lack of guidance on effective IPV prevention in Bangladesh resulted in targeting only women in most of the programmes. METHODS: This paper assesses impact of SAFE, a 20-month intervention (March 2012 to October 2013) in slums of Dhaka on IPV and tests effectiveness of female only groups vs. no groups; and female + male groups vs. female only groups on IPV in the community using a three-arm cluster randomized controlled trial. SAFE's core activities included interactive group sessions, community mobilisation, and services. The last two activities were common across arms. FINDINGS: Regression analyses (female survey: baseline n = 2,666; endline n = 2,670) showed no effect of SAFE on IPV against women aged 15-29. However, sub-group analyses demonstrated 21% risk reduction of physical IPV against adolescent girls aged 15-19 in the female + male group intervention arm. A consistent reduction in sexual violence was observed in both female and female + male arms for both groups of women, but the results were not statistically significant. INTERPRETATION: The findings emphasise the importance of combining male and female interventions for reducing physical IPV against adolescent girls. Implications for future research have been discussed.