Research That Matters (January 17 - 20, 2008)

Purpose: Although residential care placements are typically considered a last resort for children with emotional and behavioral problems in the child welfare system, limited information is available on the extent and predictors of residential care placement among children and adolescents with mental health needs. This study longitudinally followed a cohort of children in out-of-home care who experienced mental health crisis in order to observe the extent of subsequent placement in residential care among the children and to examine the roles of crisis assessment, type of treatment, and experiences in out-of-home care on residential care placement.

Method: The analyses were conducted using child welfare records and the Screening, Assessment, and Supportive Services (SASS) program files from the state of Illinois. All children in protective custody in Illinois are eligible for the SASS program. The sample comprised 685 children and adolescents between the ages of 3 and 21 years who were in non-residential care out-of-home placement and who were referred to the SASS program for emotional and behavioral problems for the first time between 2002 and 2003. As there were children in recurrent psychiatric crisis, a total of 1,118 SASS episodes of 685 children were included for analyses. Residential care was measured as being placed in group homes, residential treatment centers, and other congregate care. Mental health needs were measured during the crisis assessment via a standardized assessment tool with a 27-item Likert-type rating scale for five impairment categories of psychiatric symptoms, risk behaviors, level of functioning, co-morbidity, and system factors. Type of treatment was categorized as hospitalization and community-based treatment. Chi-square tests and ANOVA were used to explore the relationship between residential care placement and children's clinical and nonclinical characteristics. Cox regression with time-dependent variables was applied to examine the effects of covariates on the occurrence and timing of residential care placement.

Results: Approximately 36% of the sample was placed in residential care subsequent to a mental health crisis. Risk behaviors rather than psychiatric symptoms played a significant role in children's entry into residential care. Children from kinship care were at great risk for residential care compared to those from regular foster care, although clinical profiles in both groups were quite similar. Children from treatment foster care were also at great risk for entry into residential care. Other significant factors included older age, minority status, and caregivers' capacity to care for children.

Implications: This study confirms that a considerable proportion of children in mental health crisis end up in residential care afterward. The findings identify risk factors for residential care placement, which can help mental health and child welfare service providers use their limited resources for those who need timely and integrated child welfare and healthcare services. The results also suggest that not only children with mental health problems but also their foster families may benefit from continued follow-up and supportive services. Costs associated with these interventions would be offset by obviating children's placement in residential care that is associated with considerable public costs.