The Use of Text Messaging as an Adjunct to Internet-Based Cognitive Behavioral Therapy for Major Depressive Disorder in Youth: Secondary Analysis

Abstract

BackgroundAs an established treatment for major depressive disorder (MDD), cognitive behavioral therapy (CBT) is now implemented and assessed in internet-based formats that, when combined with smartphone apps, enable secure text messaging. As an adjunct to such internet-based CBT (ICBT) approaches, text messaging has been associated with increased adherence and therapeutic alliance. ObjectiveThis study analyzed data from the intervention arm of a randomized control trial evaluating 24-week ICBT for MDD (intervention arm) against standard-care psychiatry (waitlist control). The aim of this secondary analysis was to assess MDD symptom improvement in relation to the frequency and content of text messages sent by ICBT participants to Navigator-Coaches during randomized control trial participation. Higher text frequency in general and in 3 conceptual categories (appreciating alliance, alliance building disclosures, and agreement confirmation) was hypothesized to predict larger MDD symptom improvement. MethodsParticipants were young adults (18-30 years) from the Centre for Addiction and Mental Health. The frequencies of categorized texts from 20 ICBT completers were analyzed with respect to MDD symptom improvement using linear regression models. Texts were coded by 2 independent coders and categorized using content analysis. MDD symptoms were measured using the Beck Depression Inventory-II (BDI-II). ResultsParticipants sent an average of 136 text messages. Analyses indicated that BDI-II improvement was negatively associated with text messaging frequency in general (β=–0.029, 95% CI –0.11 to 0.048) and in each of the 3 categories: appreciating alliance (β=–0.096, 95% CI –0.80 to 0.61), alliance building disclosures (β=–0.098, 95% CI –0.28 to 0.084), and agreement confirmation (β=–0.076, 95% CI –0.40 to 0.25). Altogether, the effect of text messaging on BDI-II improvement was uniformly negative across statistical models. More text messaging appeared associated with less MDD symptom improvement. ConclusionsThe hypothesized positive associations between conceptually categorized text messages and MDD symptom improvement were not supported in this study. Instead, more text messaging appeared to indicate less treatment benefit. Future studies with larger samples are needed to discern the optimal use of text messaging in ICBT approaches using adjunctive modes of communication. Trial RegistrationClinical Trials.gov NCT03406052; https://www.clinicaltrials.gov/ct2/show/NCT03406052