Evidence for the Escalation of Domestic Violence in 911 Call Records

Abstract

This study was a population-based, retrospective, cohort study that examined the trajectory of domestic violence within individual households as reported via emergency 911 calls. A contextual, multi-level, geographically-referenced model was used to explain the relationships among neighborhood level factors, social disorganization indicators; household factors, number of previous episodes and total number of calls; and the escalation of domestic violence, the interval between calls and the level of severity of episodes, within individual households. The neighborhood data were derived from 1990 U.S. census data. Data on the number of domestic violence calls from individual households, the interval between each call, and the level of severity of the episode were derived from the 1997 emergency 911 database of a large southeastern city in the U.S. The pattern of recurrent calls from individual households was examined. Each emergency 911 call in the sample was coded for a set of variables. Associations among neighborhood variables that included economic status, family structure, racial composition, residential mobility, and structural density and the two outcome measures were examined. To estimate the effects of the neighborhood variables, the emergency 911 calls were linked to their respective census tracts using street addresses. Data were analyzed using a hierarchical approach. Evidence was found for the escalation of domestic violence. The number of days between consecutive episodes of violence decreased as the number of episodes of violence within a household increased. The severity of episodes of violence also increased with each subsequent episode of reported violence from the household. The proportion of female-headed householders with children in a neighborhood, a measure of family structure, was related to both the initial call interval and severity of the initial episodes of violence in households. There was significant unexplained variation among households within neighborhoods for both outcome measures. Further, neighborhood social disorganization was related to the rate of domestic violence in neighborhoods. Emergency 911 call data could provide a promising source of data for a domestic violence surveillance system, especially when linked to other data sources such as official police records.