Longitudinal Modeling of Mental Health Need and Mental Health Service Utilization Among Youth in Foster Care

Background and Purpose: Youth in foster care have higher than normative need for mental health services (MHS), and use MHS more frequently than other youth. Extant studies, often drawn from the National Survey of Child and Adolescent Wellbeing, suggest that need for MHS and MHS usage rates covary strongly (Burns et al., 2004; Horwitz et al., 2012). Despite research suggesting that need for MHS is time-variant (Villodas et al., 2014), few studies have tested whether the relationship between need for MHS and MHS utilization changes over time. The current study undertook an examination of this time-dependent relationship between MHS need and MHS use for a sample of 328 adolescent foster youth in a metropolitan Oregon region. 

Methods: Quantitative data were drawn from the Supporting Siblings in Foster Care study, a longitudinal experimental test of a sibling relational enhancement intervention (McBeath et al., 2014). Youth ranged from 7-15 years old at study enrollment; over 60% of participants were youth of color. At baseline, participants completed the Child Behavioral Checklist (CBCL) internalizing and externalizing subscales, and reported their gender, race/ethnicity, and age. At each of 4 data collection waves (spaced every 6 months), participants completed the CBCL and reported whether they had received any MHS from a professional provider. MHS utilization variables were dichotomized, and used as dependent variables in a multilevel mixed-effects regression model, to examine change in MHS utilization patterns over a total of 180 days.

Results: At baseline, the average score on the CBCL internalizing and externalizing subscales was 57.6 (range 33-86) and 59.8 (range 33-95), respectively. About 60% of youth reported mental health service utilization at any point throughout study. Bivariate analyses showed that higher baseline CBCL internalizing and externalizing scores were significantly associated with being male. The mixed-effects regression model showed that controlling for gender, race and age, respondents with higher CBCL internalizing scores were equally likely to utilize MHS at baseline, but showed higher odds of service utilization over time, compared to respondents with lower CBCL internalizing scores. However, respondents with higher CBCL externalizing scores were significantly more likely to utilize MHS at baseline, but over time, were no different than respondents with lower CBCL externalizing scores in terms of MHS utilization.

Conclusions and Implications: These findings suggest that the relationship between need for MHS and MHS use vary over time depending on mental health symptomatology. The current study found that odds of service use varied by externalizing or internalizing symptoms, with externalizing behavior associated with higher baseline MHS use but not increased odds of use over time, while internalizing behavior was associated with lower MHS utilization at baseline with some evidence of increased odds of use by time. These findings highlight the importance of research differentiating between the symptom profiles of youth with high overall MHS needs, as well as the value of examining whether these profiles change over time in relation to the provision of appropriate MHS. Findings also highlight the importance of ongoing assessment of MHS needs among foster youth.