"I never expected that it would happen, coming to ask me such questions":Ethical aspects of asking children about violence in resource poor settings

International epidemiological research into violence against children is increasing in scope and frequency, but little has been written about practical management of the ethical aspects of conducting such research in low and middle-income countries. In this paper, we describe our study procedures and reflect on our experiences conducting a survey of more than 3,700 primary school children in Uganda as part of the Good Schools Study, a cluster randomised controlled trial of a school-based violence prevention intervention. Children were questioned extensively about their experiences of physical, sexual, and emotional violence from a range of different perpetrators. We describe our sensitisation and consent procedures, developed based on our previous research experience and requirements for our study setting. To respond to disclosures of abuse that occurred during our survey, we describe a referral algorithm developed in conjunction with local services. We then describe our experience of actually implementing these procedures in our 2012 survey, based on reflections of the research team. Drawing on 40 qualitative interviews, we describe children's experiences of participating in the survey and of being referred to local child protection services. Although we were able to implement much of our protocol in a straightforward manner, we also encountered major challenges in relation to the response of local services to children's disclosures of violence. The research team had to intervene to ensure that children were provided with appropriate support and that our ethical obligations were met. In resource poor settings, finding local services that can provide appropriate support for children may be challenging, and researchers need to have concrete plans and back-up plans in place to ensure that obligations can be met. The merits of mandatory reporting of children's disclosures to local services need to be considered on a case by case basis-in some places this has the potential to do harm. Research teams also must agree on what level of ancillary care will be provided, and budget accordingly. Further practical examples of how to address the challenges encountered in this work are needed, in order to build a consensus on best practices. NCT01678846 (clinicaltrials.gov), August 24, 2012.