JMIR Publications

ABSTRACT

Background:

Increasingly, there is a focus on how engagement with Digital Mental Health Interventions (DMHIs) relates to intervention effectiveness. Engagement is a multifaceted construct with objective (e.g., behavioral) and subjective (e.g., experience) components. DMHI trials have overwhelmingly focused on measuring objective engagement, and empirical research on the clinical relevance of objective or subjective engagement has yielded mixed to positive results. The few studies that have studied the clinical relevance of engagement as a multifaceted construct have yielded mixed results.

Objective:

This project expands upon this literature, examining the association between both objective and subjective engagement and clinical outcomes during a randomized controlled trial of HabitWorks, a transdiagnostic Smartphone app targeting interpretation bias.

Methods:

We conducted an exploratory factor analysis (EFA) using passively collected indicators of objective engagement and self-reported indicators of subjective engagement. We used regression to understand the unique relation between each factor and change in anxiety, depression, and functional impairment.

Results:

Our EFA yielded four factors, Prompted Use, Self-Initiated Use, Experience, and Usability. We found that Experience, which included affective and cognitive indicators of experience with HabitWorks, predicted improvement in anxiety, depression, and functional impairment.

Conclusions:

Our factor structure of engagement aligns with the literature's typical distinction between objective and subjective engagement but suggests that within these domains there may be meaningful divisions that denote unique underlying processes. Findings further indicate that experience with the app is clinically meaningful, suggesting that developer and clinical scientist's attention to this construct is critical to reducing user burden while increasing intervention effectiveness. Clinical Trial: This study is secondary data analysis of a recent randomized trial and was pre-registered (10.17605/OSF.IO/SKFPG)