Respondent-driven sampling to assess outcomes of sexual violence: a methodological assessment.

Abstract

Sexual violence is pervasive in eastern Democratic Republic of Congo (DRC). Survivors of sexual violence encounter numerous challenges, and women with a sexual violence-related pregnancy (SVRP) face even more complex sequelae. Because of the stigma associated with SVRP, there is no conventional sampling frame and, therefore, a paucity of research on SVRP outcomes. Respondent-driven sampling (RDS), used to study this "hidden" population, uses a peer recruitment sampling system that maintains strict participant privacy and controls and tracks recruitment. If RDS assumptions are met and the sample attains equilibrium, sample weights to correct for biases associated with traditional chain referral sampling can be calculated. Questionnaires were administered to female participants who were raising a child from a SVRP and/or who terminated a SVRP. A total of 852 participants were recruited from October 9, 2012, to November 7, 2012. There was rapid recruitment, and there were long referral chains. The majority of the variables reached equilibrium; thus, trends established in the sample population reflected the target population's trends. To our knowledge, this is the first study to use RDS to study outcomes of sexual violence. RDS was successfully applied to this population and context and should be considered as a sampling methodology in future sexual violence research.