OBJECTIVE: This study compared recidivism rates for alleged maltreatment perpetrators whose initial report was substantiated and those whose initial report was not substantiated, to determine whether they returned to the child welfare system at differing rates. METHOD: Statewide administrative child welfare services data over a period of 4 1/2 years was analyzed. The sample included 31,531 perpetrators of intrafamilial maltreatment. Separate analyses were conducted for each type of maltreatment (sexual abuse, physical abuse, and neglect). RESULTS: Bivariate relationships (substantiation status and recidivism) were analyzed using survival curves, and the study found that recidivism patterns differed by type of maltreatment and by substantiation status at the index event. Multivariate analyses using Cox Proportional Hazards models found that bivariate relationships held true even when controlling for neighborhood mean income, ethnicity, and gender. CONCLUSIONS: The study's finding that perpetrators whose index event was not substantiated return to the child welfare system at a high rate suggests the need to learn how to discriminate those unsubstantiated alleged perpetrators who are most at risk for recidivism. The finding of frequent cross-type recidivism indicates that intervention should focus on common themes across maltreatment types.
OBJECTIVE: This study compared recidivism rates for alleged maltreatment perpetrators whose initial report was substantiated and those whose initial report was not substantiated, to determine whether they returned to the child welfare system at differing rates. METHOD: Statewide administrative child welfare services data over a period of 4 1/2 years was analyzed. The sample included 31,531 perpetrators of intrafamilial maltreatment. Separate analyses were conducted for each type of maltreatment (sexual abuse, physical abuse, and neglect). RESULTS: Bivariate relationships (substantiation status and recidivism) were analyzed using survival curves, and the study found that recidivism patterns differed by type of maltreatment and by substantiation status at the index event. Multivariate analyses using Cox Proportional Hazards models found that bivariate relationships held true even when controlling for neighborhood mean income, ethnicity, and gender. CONCLUSIONS: The study's finding that perpetrators whose index event was not substantiated return to the child welfare system at a high rate suggests the need to learn how to discriminate those unsubstantiated alleged perpetrators who are most at risk for recidivism. The finding of frequent cross-type recidivism indicates that intervention should focus on common themes across maltreatment types.
Authors: Laura J Proctor; Gregory A Aarons; Howard Dubowitz; Diana J English; Terri Lewis; Richard Thompson; Jon M Hussey; Alan J Litrownik; Scott C Roesch Journal: Child Maltreat Date: 2012-06-20