Can Patient Characteristics at Intake Predict Patient Response to Therapist-Assisted, Transdiagnostic Internet-Delivered Cognitive Behavioural Therapy?

Abstract

Internet-delivered interventions for anxiety and depression show promise in both their effectiveness at reducing symptoms and their efficiency in routine practice. Still, some patients do not benefit from these programs, and there is limited understanding of factors that predict patient response. The current study investigates predictors of patient response to the Wellbeing Course, a transdiagnostic, therapist-assisted internet-delivered cognitive behavioural therapy designed to treat depression and anxiety. Case files from 1201 patients who participated in a recent open trial of the Wellbeing Course were analyzed to examine predictors of dropout (i.e., withdrawing from the course before completing all modules) and symptom change, which was measured weekly using the Patient Health Questionnaire-9, a measure of depressive symptoms, and the Generalized Anxiety Disorder-7, a measure of general symptoms of anxiety. Logistic regression identified younger age and higher initial scores of psychological distress as significant predictors of dropout. Autoregressive latent trajectory modelling was used to assess the value of various patient characteristics as measured at screening for predicting symptom trajectories over the course of treatment. Patients who reported being on disability and those who do not hold a university degree were found to have higher initial symptoms and experience greater reductions in symptoms over treatment. Cases were then classified as either unsuccessful (dropout or no significant symptom reduction) or successful (course completion and significant symptom reduction). Using these criteria for success as an outcome criterion, a clinical decision tree was created to guide program referrals and help therapists assess risk. Future research directions are discussed.