Violence in childhood and community contexts: a multi-level model of factors associated with women's intimate partner violence experience in Samoa

Background Intimate partner violence (IPV) affects approximately 26% of women worldwide and is driven by a complex interplay of factors across individual, relationship/household, community and societal levels. Individual and relationship/household factors are well studied however little empirical evidence exists on factors at the community level that drive IPV which are needed to inform prevention interventions. Methods We conducted a cross-sectional, multi-level analysis of factors associated with women's IPV experience in Samoa using the 2019-20 Demographic and Health Multiple Indicator Cluster Survey. We used hierarchical multivariable logistic regression to assess individual, relationship/household and community level effects on women's risk of physical, sexual and/or emotional IPV. Findings The past year prevalence of physical, sexual and/or emotional IPV among women in Samoa was 31.4%. At the individual and relationship/household level, women's employment, witnessing IPV between parents, experiencing physical abuse from a parent, and partner's alcohol use and controlling behaviours were associated with higher risk of IPV. At the community level, higher levels of women with higher education and involved in household decision-making, and higher levels of men in employment were protective against IPV. Interpretation A complex interplay of factors across individual, relationship/household and community levels are asso-ciated with women's experience of IPV in Samoa. Experiences of IPV are embedded within a broader context of violence against children and harmful alcohol use. Community contexts, including women's empowerment and men's employment, are also associated with women's IPV experience in Samoa. These findings not only demonstrate that public health issues such as IPV, violence against children and harmful alcohol use should be addressed together as part of multipronged approaches, but they point towards the importance of community-level analyses for designing and delivering community-based interventions. Greater knowledge of community dynamics will enable community-based interventions to create environments at the community level that support meaningful and sustainable change towards IPV prevention.