Student mental health outcomes of a clustered SMART for developing an adaptive implementation strategy to support school-based CBT delivery

Purpose: Most youth experiencing anxiety/depression lack access to evidence-based mental health practices (EBPs). School-delivered care improves access, and various support can help school professionals (SPs; school social workers, counselors) deliver EBPs, like Cognitive Behavioral Therapy (CBT). Understanding implementation strategies' impact on downstream mental health outcomes is crucial to scaling up EBPs to address the treatment gap, but it has rarely been assessed. Methods: This paper compares implementation strategies' impact on change in student outcomes, collected as exploratory outcomes from a type III hybrid implementation-effectiveness trial. A clustered, sequential, multiple- assignment randomized trial design was used, which embedded four implementation supports that differentially sequence three implementation strategies, Replicating Effective Programs (REP), Coaching, and Facilitation. Prior to the first randomization, N = 169 SPs from 94 Michigan high schools each identified up to 10 students whom they believed could benefit from CBT and facilitated student survey completion. Changes in students' depression (Patient Health Questionnaire-9, modified for teens) and anxiety symptoms (Generalized Anxiety Disorder-7) over 10 months were compared across the four sequences of implementation support using a generalization of a marginal, weighted least squares approach developed for a clustered SMARTs. Results: Small, non-clinically significant reductions in symptoms over the study period were found. Pairwise comparisons found no significant differences in symptom change across the four implementation strategies. The difference in the estimated mean PHQ-9T/GAD-7 scores between the least and the most intensive strategies (REP vs. REP+Coaching+Facilitation) was 1.04 (95%CI =-0.95, 3.04) for depression and 0.82 (95%CI =-0.89, 2.52) for anxiety. Discussion: No difference in symptom change was found across the four implementation strategies. Multiple forms of implementation support may be useful for improving student mental health outcomes.