Hugh McCall

Does persuasive design predict efficacy in unguided ICBT? A meta-regression analysis

Masters thesis results, Hugh McCall 2020 

Background: Internet-delivered cognitive behavioural therapy (ICBT) is an innovative treatment for common mental health problems like depression and anxiety. It is similarly efficacious to face-to-face therapy when it is offered with therapist support by phone or secure email. It is not as efficacious when it is offered in a purely self-guided format, but it is easier to implement on a large scale. Various research groups have suggested that persuasive design—designing an intervention to be more engaging and stimulating for users—can help improve the efficacy of ICBT and other digital interventions for health and mental health problems. The objective of this study was to explore the relationship between persuasive design elements and efficacy among previously published trials of unguided ICBT for depression and anxiety.

Method: A systematic review of five databases was conducted to identify randomized controlled trials of unguided ICBT for depression and anxiety. Meta-analyses were conducted, and meta-regression was used to determine whether the effect size of each unguided ICBT program could be predicted by the number of persuasive design elements it included. Meta-regression was also used to explore changes in the efficacy and persuasive design of unguided ICBT over time.

Results: A total of 41 studies (= 10,301) were identified through the systematic search. Meta-analyses revealed a small to moderate weighted mean effect size (Hedges’ g = 0.29) for unguided ICBT for depression and a moderate effect size (Hedges’ g = 0.48) for unguided ICBT for anxiety. The results of the meta-regressions showed that the number of persuasive design elements identified in each intervention predicted the efficacy of ICBT for depression but not ICBT for anxiety. Additional meta-regressions showed that year of study did not predict the efficacy of each intervention or the number of persuasive design elements each intervention included.

Implications: The results suggest that more persuasively designed ICBT interventions are more efficacious, at least for treating symptoms of depression. There was no evidence that efficacy or persuasive design has increased over time in unguided ICBT for depression and anxiety. However, further research will be required to clarify the role of persuasive design in ICBT; in particular, experimental research comparing versions an ICBT intervention with and without specific persuasive design elements and qualitative research exploring how clients experience persuasive design elements would help advance this literature.