Drinking Context-Specific Dose-Response Models of Intimate Partner Violence Among an Urban Emergency Department Sample.

objectiveThis cross-sectional study evaluates the association between physical intimate partner violence (IPV) and frequencies of drinking and volume consumed in six different contexts among a sample of urban emergency department (ED) patients.

methodWe obtained survey data from 1,037 married, cohabiting, or partnered patients (53% female; 50% Hispanic; 29% African American) at a Northern California safety-net hospital. Past-year physical IPV was measured with the Revised Conflict Tactics Scale. We asked patients about frequency of drinking and usual number of drinks consumed at bars, restaurants, homes of friends or relatives, own home, public places such as street corners or parking lots, and community centers or large events. Gender-stratified dose-response models were estimated for frequencies of IPV perpetration and victimization, with adjustment for sociodemographic and psychosocial factors, marijuana use, and spouse/partner problem drinking.

resultsNone of the women's context-based frequency and volume measures were associated with frequency of IPV victimization. Women's volume of alcohol consumed at home was associated positively with frequency of their IPV perpetration (β = .008, SE = .003, p < .01), and volume consumed in public places was associated negatively with this outcome (β = -.023, SE = .010, p < .05). Among men, none of the context-based frequency and volume measures were associated with frequency of either IPV outcome. Spouse/partner's problem drinking was associated with each gender's IPV victimization, and with IPV perpetration by men.

conclusionsFrequency of drinking and volume consumed in specific contexts do not substantively contribute to frequency of IPV perpetration or victimization in this sample of urban ED patients.