Novel analytic strategies are making it possible to re-examine predictors and outcomes of evidence-based treatment. Integrative data analysis (IDA) is an approach that enables the combination of multiple data sets using different measures and multiple constructs. The current study uses an IDA framework to combine existing data from eight completed randomized clinical trials of CBT for anxious youth (combined N = 687) to examine trajectories of treatment response and multilevel predictors of trajectories. Anxiety disorders are the most prevalent psychiatric conditions affecting children and adolescents, and are associated with substantial negative sequelae. The successful treatment of youth anxiety is thus a top health research priority. Results from controlled clinical trials and meta-analyses suggest that, for many youth with an anxiety disorder, cognitive-behavioral therapy (CBT) can be beneficial. However, approximately one-third of youth who receive CBT do not improve. Additionally, youth who prematurely terminate from treatment tend to show worse outcomes that those who complete a full course of treatment. One way to improve current psychotherapies for youth is to identify predictors of treatment outcomes. While research on predictors is common in the literature, results have been inconsistent. Recent recommendations outlined in the National Institutes of Mental Health Strategic Plan state that funded projects should identify individual factors that impact individual response to interventions. In line with this, the broad goals of the current research training program are to provide training to the principal investigator in the phenomenology and treatment of child and adolescent anxiety, and in state- of-the-art statistical techniques for analyzing longitudinal multi-site data. Trajectories of treatment response and multilevel predictors of trajectories will be examined across eight trials of CBT. The specific study aims are to: 1) develop a set of measures that share a standard metric across multiple studies using an item response theory (IRT) model, 2) examine patterns of treatment response by modeling trajectories of symptom change during and following CBT for anxiety disordered youth using the IRT-derived scores, and 3) examine the relationships between pretreatment youth- and treatment-specific predictor variables, early termination, and trajectories of treatment response and non-response. Data will be analyzed using IRT and latent growth mixture modeling procedures. Results from this study may lead to more personalized interventions by identifying specific youth and treatment factors associated with treatment success and failure. Thus, this study will help to fulfill NIMH's mission to "transform the understanding and treatment of mental illnesses through basic and clinical research, paving the way for prevention, recover, and cure."
The proposed research will make use of novel methodological (integrative data analysis) and analytic (item response theory;growth mixture modeling) procedures to help clarify current understanding of predictors and patterns of treatment response for cognitive-behavioral therapy (CBT) for children and adolescents. This project will generate new knowledge about treatment non-response that may lead to more personalized interventions by identifying specific youth and treatment factors associated with treatment success and failure.