Change processes in cognitive therapy for social anxiety disorder: A comparison of face-to-face and internet-based treatment formats

Background: Compared to efficacy research, studies investigating the processes of change in psychological therapy are rare, especially for internet-based interventions. While many online therapies are based on face-to-face therapy protocols, it is unknown whether the processes of clinical improvement differ between these treatment formats. Objective: To examine candidate change processes in an online therapist-guided cognitive therapy intervention for social anxiety disorder (iCT-SAD), and compare the results to the corresponding face-to-face therapy (CT-SAD). Methods: Data from a randomised controlled trial (n = 99) were analysed using Multilevel Structural Equation Models, incorporating the temporal precedence of the process variable, and disaggregating the within- and between-subject components of the predictors. These examined three candidate change processes: self-focused attention, negative social cognitions, and depressed mood. Moderated mediation models provided an additional test of the moderating effect of treatment format. Results: Negative social cognitions and self-focused attention were supported as significant mediators of clinical improvement in both CT-SAD and iCT-SAD. Effects were of similar strength and moderated mediation was not observed. There was also evidence of cyclical relationships between social anxiety symptoms and these process variables. Depressed mood also emerged as a significant but weak mediator in CT-SAD, but not in iCT-SAD. Moderated mediation was not observed. Conclusion: The online format of therapy showed a similar pattern of change processes to face-to-face treatment, with self-focused attention and negative social cognitions mediating clinical improvement in both treatments. Efforts to improve the efficacy and efficiency of SAD interventions by targeting these factors may therefore be equally applicable to online and face-to-face interventions.