Although most child welfare placements receive a rigorous assessment and ongoing safety monitoring, it is still unclear which is safest in terms of physical abuse. Our goals for this study were to assess the relative risk of physical assault for different child welfare placements in a national sample of youth and examine that risk within placement types for children with various levels of behavioral problems. Unlike previous studies of assault prevalence, we constrained our analysis to youth self-reports of assaults within a current placement to increase validity. Data from the 2010 National Survey of Child and Adolescent Well-Being (NSCAW II) were used. The final analytic sample included children 8-18 years of age at Wave 3 interview in 2012 (n = 1,302). Physical assault was measured using youth reports of being thrown at, shoved, slapped, beat-up, or stabbed, shot, or threatened with a knife or gun. Multiple bivariate contingency table analyses using Pearson χ2 tests and ordinal logistic regressions were used. Overall, 8% of youth reported a minor assault, 9% a serious assault, and 14% any assault. Ordinal logistic regression models showed that children with more severe behavioral problems were at increased odds of more serious physical assault (odds ratio [OR] = 1.03, 95% confidence interval [CI] = [1.01, 1.05]) and youth in adoptive homes were at decreased odds of assault (OR = 0.22, 95% CI = [0.03, 0.82]). Interaction effects showed that one point increase in behavioral problems resulted in increased odds of more serious assaults only for youth reunified with biological parents (OR = 1.11, p = .01) and youth living with traditional foster parents (OR = 1.06, p = .02). Our results indicate that permanent adoptive homes are the safest child welfare placements irrespective of child characteristics, which is supportive of recent efforts to increase number of youth exiting care to adoption.
Keywords: behavioral problems; child welfare; physical abuse.