Toward an mHealth App for Early Detection of Relapse in Remitted Depressed Individuals: A Focus Group Study on User Requirements and Concerns (Preprint)

Abstract

BACKGROUND Major depressive disorder is often a recurrent condition, with a high risk of relapse for remitted depressed individuals. Early detection of relapse is critical to improve clinical outcomes. mHealth (mobile health) technologies offer new opportunities for real-time monitoring and prevention of relapse, given that user requirements of the target population are effectively implemented. OBJECTIVE This study investigated remitted depressed individuals’ user requirements and concerns for an mHealth app aimed at monitoring depressive symptoms and detecting early signs of relapse through integrating both active ecological momentary assessment (EMA) data and passive data from the user’s smartphone and smartwatch. METHODS Three focus group discussions were conducted with 17 participants who had a history of depression but were in remission at the time of the study. Prior to the focus group, participants had gained some experience with an in-house designed EMA monitoring app, prompting questions regarding their mood multiple times throughout the day. During the focus groups, feedback and insights were gathered into participants’ expectations, requirements, concerns, and attitudes toward a depression monitoring app. A thematic analysis was performed to identify recurring themes and subthemes, shedding light on the desired user experience and functionalities. RESULTS We identified five main themes. Participants highlighted (1) a need for customization settings, particularly in terms of data collection and sharing, and frequency of self-assessments. They also valued (2) positivity in the app’s design through positive reinforcement and journaling features. Additionally, participants emphasized (3) interventions to be the main motivator for adoption and long-term usage. More specifically, they wanted the app to foster self-awareness, self-reflection and insights and to offer support during deteriorations in mental health. Furthermore, participants deemed (4) transparency in data use and machine learning predictions essential for building trust. Participants required these functionalities to bear (5) the user burdens of self-monitoring. Key concerns were for passive monitoring to cause a privacy burden and for active monitoring to raise an emotional burden. CONCLUSIONS Considering the vulnerability of potential users, caution is warranted in the design of an mHealth app for depression relapse prevention. Users’ requirements for customization, positivity, interventions, and transparency must be addressed, while minimizing both the emotional and privacy burden. Carefully balancing these design elements is crucial to ensure adoption and long-term user engagement.