Intimate Partner Violence Risk Assessment in Community Health Facilities: A Multisite Longitudinal Study

Abstract

The remarkable development of tools for the structured assessment of violent reoffending of intimate partner aggressors coexists with a relative scarcity of tools adapted for the diversity of community facilities for women who suffer intimate partner violence (IPV). This situation prompted the development of the RVD-Bcn in Catalonia (Spain). The present multisite prospective study is the first to test the predictive validity of this tool for IPV risk assessment in a sample different to the original. Trained assessors interviewed 321 women in five community health facilities in Cordoba (Argentina); 230 of them completed a follow-up that ranged from 47 to 311 days. Incidents of physical IPV, including serious threats and forced sex, were recorded prospectively. Rates of IPV in the low, moderate and high-risk groups were 10, 57, and 33 percent, respectively. The summary risk ratings were significantly associated with the occurrence of violence during follow-up, as were RVD scores (ROC-AUCs = .88 and .89, respectively). Low-risk participants not only suffered fewer IPV events than participants in the moderate and high-risk group, but also suffered them later in time. The most accurate predictive model of IPV included total scores and a dichotomised variable of low vs. moderate-high summary risk rating. Findings contribute to the emerging practice of including structured violence risk assessment in Latin American professional contexts and institutions.