(PS3-12) How many anxious kids in community mental health would be eligible for an RCT? And does it matter?

Abstract Text:

Background: Decades of randomized controlled trials (RCTs) provide robust efficacy data supporting cognitive-behavioral therapy (CBT) for the treatment of pediatric anxiety disorders. By design, these trials employ strict exclusion criteria to scientifically isolate effects of interest. However, this raises challenges when CBT moves from efficacy to effectiveness, as routine clinical care settings often do not exclude patients based on common trial exclusion criteria (e.g., clinical comorbidities, medication use, etc.). Thus, there are questions of whether findings from tightly controlled clinical trials can translate to naturalistic settings. We examined: (1) the proportion of patients in an urban community mental health center that would have been excluded from one or more RCTs in the published literature, and (2) if such exclusion related to treatment outcomes.

Methods: We reviewed 86 RCTs from the past 25 years that evaluated the efficacy of CBT in youth populations diagnosed with anxiety disorders, OCD, or school-related anxiety and identified their most common exclusion criteria. We compared exclusion criteria for these RCTs with demographic and clinical variables of a naturalistic patient sample receiving exposure-based treatment for anxiety or obsessive-compulsive disorder (OCD) within a specialty treatment program in an urban community health center in the Northeastern United States. Data were drawn from a chart abstraction of three years of data for consecutively presenting, racially and economically diverse youth (n = 94; M age = 12.38; 66.7% female; 56.3% White, 14.6% Black, 10.5% Asian, 7.3% Mixed Race; 10.5% Hispanic/Latine and White).

Results: The most common exclusion criteria across RCTs were (a) depressive disorder, (b) attention-deficit/hyperactivity disorder (ADHD), (c) autism spectrum disorder, (d) suicidality, (e) psychosis, and (f) current medication use. Based on these criteria, 53% (n = 51/96) of participants in our sample would have been excluded from one or more RCTs. There was a significant association between participant exclusion from RCTs and successful treatment graduation. Patients excluded from one or more RCTs had 2.62 greater odds of not fully terminating treatment successfully relative to those who met RCT inclusion criteria (OR = 2.62, 95% CI [1.14, 6.06], p = .02). However, youth were equally likely to meet anxiety specific goals and graduate from anxiety specialty treatment (OR = 1.55, 95% CI [0.63, 3.80], p = .34).

Conclusion: Over half of youth treated in this community setting would have been excluded from one or more RCTs. These findings suggest youth who did not match RCT inclusion criteria had poorer overall treatment prognosis. Implications for future treatment research will be discussed.