A comparison of two group cognitive behavioral therapy protocols for anxiety in urban schools: appropriateness, child outcomes, and cost-effectiveness.

Background: Cognitive behavioral therapy (CBT) for pediatric anxiety is efficacious for reducing anxiety symptoms and improving functioning, but many children are unable to access CBT for anxiety in community settings. Schools are an important setting in which children access mental health care, including therapy for anxiety. In this setting, therapy is usually delivered by Masters-level therapists. Objectives: Friends for Life (FRIENDS), a 12-session, manualized, group CBT program for anxiety has demonstrated effectiveness when implemented in schools. However, prior research has also found challenges regarding feasibility and cultural fit when delivering FRIENDS in the urban school context. To address these challenges, we adapted FRIENDS for implementation in the school setting so that it might be more feasible and culturally appropriate for low-income, urban schools in the United States, while maintaining the core components of treatment. The current study uses a mixed-method approach to compare the effectiveness, cost-effectiveness, and perceived appropriateness of FRIENDS and CATS when delivered by Masters-level therapists with train-the-trainer support. Materials and methods: First, we compared change scores for student outcomes (i.e., child-report MASC-2 total score, parent-report MASC-2 total score, teacher-report Engagement and Disaffection subscale scores) from pre- to post- treatment between students receiving FRIENDS and students receiving CATS to assess whether the two conditions resulted in equivalent outcomes. Second, we compared the cost and cost-effectiveness between the groups. Finally, we used an applied thematic analysis to compare appropriateness of the interventions as perceived by therapists and supervisors. Results: The mean change score for the child-reported MASC-2 was 1.9 (SE = 1.72) points in the FRIENDS condition and 2.9 (SE = 1.73) points in the CATS condition; results indicated that the conditions were similar in their treatment effects, and symptom reductions were small in both groups. The modified protocol, CATS, was shown to cost significantly less to implement compared to FRIENDS and showed greater cost-effectiveness. Finally, compared to therapists and supervisors in the CATS condition, therapists and supervisors in the FRIENDS condition more strongly described aspects of the intervention that were not appropriate for their context and in need of more extensive adaptations. Conclusion: Relatively brief, group CBT for anxiety, with adaptations to improve cultural fit, is a promising approach to treat youth anxiety symptom when delivered by school-based therapists with train-the-trainer implementation support.