A Comparison of Two School Based Interventions to Manage Anxiety in Autistic Students

Background and Significance: Anxiety is very common for autistic youth and can significantly limit functioning and full participation across settings. In school, anxiety may negatively impact students' attendance, school performance, peer relationships and extracurricular participation. Unfortunately, autistic youth have experienced long-standing difficulty accessing mental health care in community settings, and this difficulty is even more pronounced for youth from historically underserved racial and ethnic backgrounds. Cognitive-behavioral therapies (CBT) are evidence-based practices that are effective for anxious children with and without autism but have not yet been consistently available in schools. Schools may be the best location to equitably manage the mental health symptoms of autistic youth. In fact, youth from low-income households and/or other underserved communities are most likely to access services only in school settings. Thus, there is a critical need for evidence-based mental health interventions in schools for autistic students. The primary purpose of the current study is to compare two school-based curriculums: Facing Your Fears - Schools Based (FYF-SB) and Zones of Regulation (ZOR) on meaningful outcomes. The results of this study will not only fill a significant gap in the literature but will provide practical information about the strengths and weaknesses of each program so that school leaders and other stakeholders can make informed decisions about program selection for autistic students.

Study Aims. Aim 1: Compare the effectiveness of FYF-SB and ZOR for 200 school-aged autistic students with interfering anxiety.

Compare the effectiveness of FYF-SB and ZOR on: (a) symptoms of anxiety (primary) and emotion dysregulation (co-primary) as rated by caregivers, students, teachers, and masked research staff; and (b) functional outcomes of priority to stakeholders, including school attendance, disciplinary action, and academic participation (secondary). Hypothesis: Students who received FYF-SB will show greater reduction of anxiety compared to students who received ZOR.

Aim 2. Compare the acceptability, appropriateness, and feasibility of FYF-SB and ZOR via a mixed methods approach with students, caregivers and school providers to aid in decision-making regarding program selection (Secondary). Following intervention completion, participants will complete measures on acceptability, appropriateness, and feasibility of either FYF-SB or ZOR. In addition, a purposive subset of participants will participate in exit interviews to understand outcomes of interest, including potential harms linked with either FYF-SB or ZOR.

Aim 3. Examine autism symptom severity and anxiety symptoms severity as moderators of treatment response. This exploratory aim may help answer questions regarding which program could work best for which students.

Study Description: This study is a multi-site cluster randomized controlled, type 1 hybrid comparative effectiveness-implementation trial designed to examine whether FYF-SB or ZOR is more effective in treating anxiety in autistic students. School administrators in Colorado and North Carolina will select schools with large numbers of students from historically underserved communities. Randomization will occur at the school level to either FYF-SB (50 schools) or ZOR (50 schools). Interdisciplinary school providers (ISPs) will be trained in one of the interventions and then nominate autistic students with anxiety for participation. ISPs will deliver the intervention programs to small groups (2-5 students) in their schools. Assessments will be conducted pre, post and 6 months following program completion. Independent evaluators masked to condition will complete post-assessments. Implementation outcomes (acceptability and feasibility) will also be examined via mixed-methods and reported by students, caregivers and ISPs.

Main components of the intervention and comparator (s): The two school-based curriculums in this study are: Facing Your Fears- School Based (FYF-SB) and Zones of Regulation (ZOR). Both are rooted in CBT principles but manage anxiety in different ways. FYF-SB is based on disorder-specific CBT strategies tailored for reduction of anxiety symptoms and incorporates psychoeducation and graded exposure (facing fears a little at a time) to target anxiety. ZOR is an example of a transdiagnostic approach to intervention and targets emotion dysregulation by focusing on four different emotional states called zones. Students are taught strategies for managing emotions and how to stay in a certain zone. A transdiagnostic treatment approach posit that when emotion regulation improves so to will psychiatric symptoms related to those emotions (e.g., anxiety).

Study Population: There are three groups of participants in the study: autistic students, caregivers and interdisciplinary school providers (ISPs). (1) 200 autistic students with interfering anxiety (8-14 years) from historically underserved communities attending elementary or middle schools in Colorado or North Carolina; (2) 200 caregivers of the eligible autistic students; and (3) 150 ISPs (e.g., school psychologists, special educators, speech/language pathologists, occupational and physical therapists) who work with autistic students

Primary Outcomes: The primary outcome (s) for this study are: (1) Screening for Childhood Anxiety and Related Emotional Disorders - Child/Parent versions (SCARED; Birmaher et al. 1999). The SCARED is a 41-item inventory of statements that assesses five types of anxiety experienced by children and adolescents, to be completed separately by caregivers and students pre/post intervention and at 6 months follow-up. A total score, as well as cutoffs for specific domains of anxiety (e.g., social, generalized) are obtained. (2) Co-Primary Outcome: Pediatric Anxiety Rating Scale- Autism Spectrum Disorder (PARS-ASD) (Maddox et al. 2020). The PARS-ASD is a clinician-rated semi-structured interview assessing anxiety severity and impairment over the past week, modified for youth with ASD ages 5 to 17 from the original Pediatric Anxiety Rating Scale (PARS) (RUPP) (Walkup et al. 2013). The PARS-ASD will be administered by an independent evaluator (IE) masked to condition. It includes: a Symptom Checklist and five Severity Items.

Secondary Outcomes: There are multiple secondary outcomes: (1) stakeholder (patient-centered) identified meaningful outcomes (e.g., tardies, school attendance, grades, number of elopements, classroom participation, use of coping strategies, increase in social initiations, number of phone calls home, visits to the principal, and number of expulsions/suspensions); (2) Emotion Dysregulation Inventory (EDI) (Mazefsky et al. 2018). The EDI is a parent report measure of emotion dysregulation in youth with ASD and will be completed pre/post intervention and at follow-up. The EDI can be used as a self-report measure and can also be completed by school providers); (3) The School Anxiety Scale-Teacher Report (SAS-TR) (Lyneham et al. 2008) is a 16-item teacher-reported measure of anxiety designed to assess the behavior of children at school from 5 to 12 years of age. Items are answered on a four-point scale. The measure provides a total score for anxiety (scores ranging from 0-48); (4) Strengths and Difficulties Questionnaire (SDQ) (Goodman et al. 2000) is a 25-item teacher, parent and student reported tool measuring prosocial behavior and psychopathology of youth 3-16 years old;(5) Emotion Regulation Index for Children and Adolescents (ERICA) (MacDermott et al. 2010) is a 16-item self-report measure of social competence and emotion regulation for children between the ages of 9-16.

Implementation Outcomes: (1) CBT Program Use. This brief measure will be administered to ISPs to obtain information about their knowledge and perceptions related to school based interventions, including ZOR and FYF-SB. (2) Acceptability Questionnaire. A six-item acceptability questionnaire will be completed by students and caregivers asking them to respond to questions such as "How much did you enjoy participating in FYF-SB/ZOR?" or "My child is less anxious after participating in FYF-SB/ZOR?" A 5-point Likert scale will be used to rate their responses. (3) Implementation Survey: Brief evaluation forms using a 6-point Likert scale will be completed by ISPs post delivery of FYF-SB/ZOR to assess quality of the training, feasibility, acceptability, and appropriateness of each program including the likelihood that providers will continue to implement the interventions once the study has ended; (4) Semi-structured Exit Interviews: A purposive subset of participants (students, ISPs, caregivers) will be engaged in semi-structured exit interviews conducted by an independent member of the research team designed to understand the acceptability, appropriateness, feasibility and potential harms or negative aspects of either FYF-SB and ZOR. (5) FYF-SB/ZOR Treatment Fidelity Checklists (for internal use only). These checklists will assess ISPs' adherence to either core components of FYF-SB or ZOR, including the four ZOR key elements of preparation, provision of structure, quality of facilitation, and adult evaluation of student learning. The measures assesses the presence/absence of core treatment components as well as quality of intervention delivery.

Timeframe: Assessments occur pre and post-intervention for primary and secondary outcomes. Selected measures will also be administered at 6 months follow-up, with the 6 month follow up as primary endpoint.

Analytic Plan:

General consideration of statistical modeling: Between-arm imbalance of baseline characteristics will be examined using t and chi-square tests and evaluated based on summary statistics. Intervention effect of FYF-SB as compared to ZOR will be assessed using Linear mixed effect model (LMM) for continuous outcomes and non-linear mixed effect model (NLMM) for categorical outcome such as school attendance. To account for design effect of clustering trials, random school (cluster) effect will be in the models for outcome without repeated measures while appropriate covariance structure of LMM/NLMM accounting for both the design effect and the correlation of repeated measures will be used for outcome with repeated measures. Transformation of outcome (ie. Square root or log) will be used if normality assumption is violated. Any imbalanced potential confounding variables will be adjusted for modeling. Primary analysis will be conducted on intent-to-treat basis (ITT). LMM and NLMM will the primary method for dealing with missing value without data imputation since missing pattern would most likely be of the missing at random (MAR) pattern. The change score from baseline to the end of intervention is considered to the primary effectiveness endpoint. No type I error adjustment for multiple tests in a single model or multiple outcomes will be applied. P < 0.05 is deemed statistically significant. No interim analysis will be performed. To examine the robustness of primary analyses, sensitivity analyses will be conducted, including the same analyses among completers and analyses after missing data were imputed using Markov chain Monte Carlo (MCMC) method.