Testing Delivery Modalities of Team-Based CBT for Pediatric Anxiety in Community Health Settings.

Anxiety disorders are a significant public health problem that cause suffering for youth and their families, derail normal childhood development, and when left untreated persist into adulthood with high societal costs. Cognitive Behavioral Treatment (CBT) including exposure therapy is the treatment of choice, yet multiple barriers limit access to care. Youth from historically marginalized groups face even greater barriers to accessing high-quality care, including limited availability and inconvenience of services, transportation difficulties, lower mental health literacy, and experiences of stigmatization and discrimination. Systematic reviews consistently show comparable outcomes for in-person vs. telehealth treatment for anxiety yet results highlight the lack of rigorous comparative research in understanding which patient subgroups are more likely to benefit from different delivery modalities, making the process of selecting among them more confusing than ever for families. In the wake of the COVID-19 pandemic, using a flexible service modality (i.e., custom mix of telehealth and in-person sessions) has also become common practice. The question of how best to improve access to high quality care is particularly relevant in community health settings (e.g., Certified Community Behavioral Health Centers (CCBHCs) and Federally Qualified Health Centers (FQHCs)) that serve a high proportion of vulnerable youth due to the staggering and longstanding failure to include these youth in clinical trials and the clear need to tailor treatment to better address barriers to access, quality, and clinical improvement. The overall goal of this proposal is to compare three modalities for delivering outpatient exposure-based CBT in youth ages 5-18 using existing care teams in CCBHC/FQHC ("safety net") settings: in-person (traditional face-to-face sessions only, occurring in the office and home/community) vs. telehealth (only occurring remotely via web-based video conference) vs. flexible (tailored mix of telehealth and/or in-person sessions). A total of 501 children with anxiety and/or OCD will be treated, with 167 children randomly assigned to each of the treatment conditions. A trained independent evaluator (IE) masked to condition will measure patient and family treatment engagement, anxiety symptoms/severity, and functional impairment at baseline, during active treatment at weeks 8, 16, and 24, and 3- and 6-month follow-up after acute treatment. The primary aims are to compare the relative effectiveness of in-person vs. telehealth vs. flexible CBT for: reducing anxiety-related functional impairment, improving family treatment engagement, improving family satisfaction, reducing barriers to treatment access, and maintaining treatment quality. Secondary outcomes are to understand predictors of pre- to post-treatment change in functional impairment including severity of illness, caregiver burden, and family accommodation and to explore group differences in time course of response and durability of treatment gains. Longer-term goals include establishing a service delivery model that can be scaled nationally in similar practice settings. Primary outcomes are driven by real world needs of patients and families and selected in collaboration with stakeholders. The approach for all aims will involve regression of the outcome variable on treatment group (flexible vs in-person or telehealth), randomization stratification factors (minoritized race/ethnicity, site), a small number of pre-specified covariates that are hypothesized to be strongly associated with the outcome (age, poverty), and baseline value of the outcome, as appropriate. This is the ANCOVA approach to analysis in randomized controlled trials. Missing data will be handled with multiple imputation using chained equations. In secondary analyses, we will evaluate the potential impact of missing data and assumptions regarding the relationship of missing data mechanisms and treatment effect estimates using sensitivity analyses.