Research That Matters (January 17 - 20, 2008)

Purpose: Substance abuse, domestic violence, and mental health are well-known risk factors for child maltreatment, and families with co-occuring problems face poor outcomes in the child welfare system (Drake et al. 2006, Marsh et al. 2006; Besinger et al 1999, Zuravin and DePanfillis, 1997). However, empirical research is limited on to what extent co-occuring problems contibute to child welfare outcomes. This study examines the effects of co-occuring problems on the likelihood of placement among the cases investigated and later sustantiated for child maltreatment by Child Protective Serivces in a rural county. The study adds to the current knowledge by investigating the role of each co-occruing problem on placement.

Method: The sample consisted of 220 families substantiated for child maltreatment in a rural county in upstate New York as part of an evaluation of collocated substance abuse servcies. First, all indicated cases in the study county were identified for two different time periods (1998-2000 and 2002 – 2004). Then a random sample was drawn with overrepresentation of cases with substance abuse issues. After the sample was identified, case records for each family in the sample were reviewed at the study site. Presence of mental health, substance abuse, and domestic violence in the case records were recorded. Socio demographic data such as mother's age, number of children, and household structure were also collected. Data on child placement was measured by any substitute care other than the current primary care provider over the 15 month period after the intial investigation. Logistic regression models were used to estimate the risk of each co-occuring problem on the likelihood of placement.

Results: The prevalence (18%) of mental health problem was lower than the prevalence of domestic violence (42%) in this sample. The hypothesized models provided a satisfying goodness of fit. Maternal mental health was a signficant predictor for child placement while domestic violence was not. Holding other variables constant, mothers with mental health problems were 2.3 times more likey to have her child placed in a substitute care (p= .03). Mother's substance abuse (p=.03) and age (p=.02) were also significantly related to child placement. Among cases identified with parental substanace abuse problem (n=136, 62% of 220), maternal mental health signifantly increases the risk of placement (AOR=2.62 p=.04).

Implications: This study confirms that families with co-occuring problems experience poor child weflare outcomes. While substance abuse and mental health problems significantly increase risk of child being placed in a substitute care, the presence of domestic violence may not. It is plausible that while domestic violence is more prevalent thatn other problems it may not directly related to mother's ability to care for the child. These findings however needs to be intrepreted with caution, given the overrepresentation of the substance abuse cases in the sample. The study does suggest that child welfare practice should addresses multiple problems in the famillies in order prevent child placement.