Predictors of foster care youths’ engagement in preventive interventions

Hambrick, E., Lee, S.-K., Weiler, L., Collins, J. O., Rhodes, T., & Taussig, H. (2022). Engagement in a preventive intervention for preadolescent children in foster care: Considerations for intervention design. Child Psychiatry & Human Development. 

https://doi.org/10.1007/s10578-022-01341-8

Summarized by Ariel Ervin

Notes of Interest:

• High levels of engagement are essential for preventative interventions. However, it’s challenging to keep children engaged.
• This study examines child, demographic, and system-level predictors of child engagement in Fostering Healthy Futures program.
• Findings indicate males were a) rated as less engaged in skill groups & mentor visits and b) participated in fewer skill groups.
• Child sex, behavioral problems, trauma symptoms, and IQ demonstrated bivariant associations with engagement & attendance.
• Group engagement correlated with fewer behavioral issues and higher IQs.
• Group attendance was negatively correlated with adverse childhood experiences (ACEs) and positively correlated with trauma symptoms.
• Programs need to identify and target youth who are more likely to engage with their interventions.

Introduction (Reprinted from the Abstract)

Engagement in mental health-focused preventive interventions is understudied. Demographic, child, and system-level predictors of engagement were explored in a study with children in foster care (N = 222, Mage = 10.3) who participated in a 30-week intervention. Attendance and engagement in mentor visits and skills groups were rated weekly. Only 4 of 21 predictors showed bivariate associations with attendance/engagement: child sex, IQ, behavior problems, and trauma symptoms. SEM models with these three variables and a measure of adverse childhood experience (ACEs), were used to develop a model of engagement. Males had poorer mentor visit and group engagement. Group attendance was positively associated with trauma symptoms and negatively associated with ACEs. Group engagement was associated with higher IQ and fewer behavior problems. A contextually-sensitive intervention can result in high engagement for a vulnerable and diverse population, yet a few child factors still impacted engagement, and when identified could be ameliorated.

Implications (Reprinted from the Discussion)

A sufficient degree of engagement in mental health-focused preventive interventions may be required to receive intended intervention benefits [1,2,3]. Little is known, however, about engagement in child focused mental health preventive interventions, particularly those serving children with unique engagement challenges such as children in foster care. This study examined demographic, child, and system-level predictors of engagement in Fostering Healthy Futures (FHF), a child-focused preventive intervention for children in foster care.

In general, there was a high degree of child engagement in FHF. Only 16 participants (~ 7%) dropped out across the 10 cohorts (years) of the program, and the average number of mentoring visits and skills groups attended (out of 30) was approximately 25. Most children had approximately 50 in-person points of connection with their mentors given that mentors both met individually with children each week and drove them to skills groups each week, and 25 points of connection with the skills group leaders/other children in the program over a 9-month period. Regarding mentor and skills group leader ratings of engagement, the average rating of the indicators of child engagement was 4.15 out of 5, suggesting that children were typically rated as quite engaged. Findings from the measurement model suggested that the model of engagement had acceptable fit when keeping the four engagement-related dependent variables separate (Fig. 1).

Given the high rates of engagement and thus the low variability in the dependent variables, it is not surprising that few significant associations between demographic, child, and system-level variables and the engagement variables emerged. Regarding demographic predictors, bivariate statistics indicated that males attended fewer skills groups and were rated as less engaged during both mentoring visits and skills groups. In the structural model (Fig. 2), male sex did not persist in predicting skills group attendance but did persist in predicting less mentoring visit and skills group engagement.

To access this article, click here.