Using Child Welfare Case Records to Understand Children's Behavioral Health Needs and Services

Background:

The behavioral health needs and extensive service gaps among children involved in the child welfare system are well documented in national surveys (Burns et al, 2004) although may vary regionally. To measure, monitor, and inform planning around children’s behavioral health services, child welfare agency administrators’ depend on locally-accessible data, and administrative data systems are a logical source (Jonson-Reid & Drake, 2008).  This study uses local child welfare administrative data to examine two aims: (1) children’s behavioral health diagnoses and service receipt as documented in child welfare administrative records, and (2) child, family, and case-level factors associated with children’s documented behavioral health service receipt.

Methods:

This study analyzes case records maintained by a public, county-based child welfare agency in the Midwest. Records were extracted for 2302 children in child welfare custody between January and June 2012. Behavioral health diagnoses entered into the database were classified into 13 major DSM-based diagnostic groups.  Behavioral health services documented within the records included counseling, consultation, and medication monitoring visits. Several other characteristics were extracted: demographics (age, race/ethnicity, gender), and child welfare case-features (abuse type, most recent placement type, abuse risk). We examined documented diagnoses, service receipt, and case characteristics using univariate and bivariate analyses (aim 1). Logistic regression was used to examine predictors of behavioral health service receipt (aim 2) for 668 (29%) children with complete case files. Based on prior findings that service receipt predictors vary by age, we ran separate models for younger (birth-10, n=278) and older children (11+, n=390).

Results:

Forty-two percent of children had a documented behavioral health diagnosis while only 37.2% of them received services; the most common diagnoses included disruptive behavior (27%), mood (19.8%), and attention disorders (17.9%).  Younger children were more likely to receive services if they experienced sexual abuse (OR=2.56, p<.05), had many children in the home (OR=2.67, p<.05), and were in non-institutional placements besides kinship or foster care (OR=3.31, p<.05). Older children were less likely to receive services if they experienced physical abuse (OR=.39, p<.05), prior maltreatment reports (OR=.60, p<.05), and strong family support system (OR=.37, p<.01). For younger and older children, African American children were less likely to receive services than whites (OR=.26, p<.05, OR=.54, p<.05, respectively). Having a mental health diagnosis was not associated with service receipt.

Conclusion:

Consistent with national estimates, this study identified gaps in behavioral health services among a local cohort of children in the child welfare system.  Having a behavioral health diagnosis had no bearing on service receipt. Rather, other demographic characteristics, placement type, and abuse risk factors drove service receipt. Results highlight concerning racial disparities and suggest that workers may be directing youth to services differently depending on their age. Our findings suggest the importance of systematic behavioral health screening and assessment within child welfare contexts to ensure that children’s clinical needs drive service receipt.  Future studies are needed that connect complete and accurate child welfare case records with behavioral health information.