Primary aim results of a clustered SMART for developing a school-level, adaptive implementation strategy to support CBT delivery at high schools in Michigan

BACKGROUND: Schools increasingly provide mental health services to students, but often lack access to implementation strategies to support school-based (and school professional [SP]) delivery of evidence-based practices. Given substantial heterogeneity in implementation barriers across schools, development of adaptive implementation strategies that guide which implementation strategies to provide to which schools and when may be necessary to support scale-up. METHODS: A clustered, sequential, multiple-assignment randomized trial (SMART) of high schools across Michigan was used to inform the development of a school-level adaptive implementation strategy for supporting SP-delivered cognitive behavioral therapy (CBT). All schools were first provided with implementation support informed by Replicating Effective Programs (REP) and then were randomized to add in-person Coaching or not (phase 1). After 8 weeks, schools were assessed for response based on SP-reported frequency of CBT delivered to students and/or barriers reported. Responder schools continued with phase 1 implementation strategies. Slower-responder schools (not providing ≥ 3 CBT components to ≥10 students or >2 organizational barriers identified) were re-randomized to add Facilitation to current support or not (phase 2). The primary aim hypothesis was that SPs at schools receiving the REP + Coaching + Facilitation adaptive implementation strategy would deliver more CBT sessions than SPs at schools receiving REP alone. Secondary aims compared four implementation strategies (Coaching vs no Coaching × Facilitation vs no Facilitation) on CBT sessions delivered, including by type (group, brief and full individual). Analyses used a marginal, weighted least squares approach developed for clustered SMARTs. RESULTS: SPs (n = 169) at 94 high schools entered the study. N = 83 schools (88%) were slower-responders after phase 1. Contrary to the primary aim hypothesis, there was no evidence of a significant difference in CBT sessions delivered between REP + Coaching + Facilitation and REP alone (111.4 vs. 121.1 average total CBT sessions; p = 0.63). In secondary analyses, the adaptive strategy that offered REP + Facilitation resulted in the highest average CBT delivery (154.1 sessions) and the non-adaptive strategy offering REP + Coaching the lowest (94.5 sessions). CONCLUSIONS: The most effective strategy in terms of average SP-reported CBT delivery is the adaptive implementation strategy that (i) begins with REP, (ii) augments with Facilitation for slower-responder schools (schools where SPs identified organizational barriers or struggled to deliver CBT), and (iii) stays the course with REP for responder schools. TRIAL REGISTRATION: ClinicalTrials.gov, NCT03541317, May 30, 2018. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13012-022-01211-w.