Social network diffusion effects of an intimate partner violence prevention program in rural Nepal

Intimate partner violence (IPV) is a pervasive global health issue. Social norms supporting male dominance are key determinants of IPV. While targeting social norms is crucial for prevention, most intervention evaluations focus only on direct beneficiaries, overlooking diffusion effects to non-participants. This study used social network analysis to assess the diffusion effects of an IPV prevention intervention in Nepal. The participants included social network members (alters) of intervention beneficiaries (egos) across 13 communities in Nawalpur District (N = 258). Data were collected at baseline and midline (at the completion of the core curriculum). Using propensity score-adjusted GEE regression models for difference-in-difference estimates, we assessed intervention effects diffusion to alters of direct beneficiaries and alters of other participants compared to control participants' alters. The results showed both types of alters experienced significant increases in positive injunctive norms compared to control alters. Similar diffusion patterns were observed for financial decision-making, relationship quality and alcohol abuse. The effects diminished in low-closeness ego-alter pairs but remained significant in high-closeness relationships. Our study suggests that diffusion is an important change mechanism in IPV interventions, particularly for normative change, though effects vary depending on the outcome type and relationship closeness between participants and their social network members.