Loading...
Using Machine Learning Methods to Predict Treatment Outcome for Anxious Youth
Norris, Lesley
Norris, Lesley
Citations
Altmetric:
Genre
Thesis/Dissertation
Date
2023
Advisor
Committee member
Group
Department
Psychology
Subject
Permanent link to this record
Collections
Research Projects
Organizational Units
Journal Issue
DOI
http://dx.doi.org/10.34944/dspace/8844
Abstract
Background: Efficacious treatments for youth anxiety disorders exist, but there is considerable heterogeneity in outcome. Predictors and moderators of differential treatment response have been difficult to identify. Machine Learning (ML) is a promising approach. Methods: Data from nine randomized controlled trials (RCTs) of youth anxiety treatments were harmonized into a dataset (N = 1362; Mage = 10.59, SDage = 2.47; 48.9% female; 71.9% White, 5.9% Black, 1.0% Asian; 10.8% Hispanic) and supervised ML algorithms predicted treatment outcomes. ML models were also built separately for youth who completed individual cognitive behavioral therapy (CBT), family CBT, combination of sertraline (SRT) and CBT, and SRT alone to examine predictive features. Models were then externally validated in a research clinic providing CBT for youth anxiety (N = 50; Mage = 12.04, SDage = 3.22; 56% female; 76% Caucasian, 10% Black, 6 % Asian, 2% Other; 6% Hispanic). Results: Lasso Regression emerged as the best performing model (RMSE = 1.40), with comparable RMSEs when the same approach was applied within an external dataset (RMSE = 1.40). Predictive features of poorer outcomes were primarily indicators of increased symptom severity, particularly youth depressive symptom severity, although predictors varied within subsamples (e.g., caregiver psychopathology was an important predictor for FCBT; increased somatic symptoms were important predictors for better response to SRT). Discussion: ML helped identify features of anxious youth who will respond to treatments.
Description
Citation
Citation to related work
Has part
ADA compliance
For Americans with Disabilities Act (ADA) accommodation, including help with reading this content, please contact scholarshare@temple.edu