Perspectives of Clinical Researchers on Engagement With Digital Mental Health Interventions: Qualitative Interview Study.

Background: While technology can widen access to mental health treatments, digital mental health interventions (DMHIs) frequently have low engagement and high dropout rates. A better understanding of user engagement with DMHIs can help researchers design technologies that users are more likely to benefit from. However, a major challenge is that the term "engagement" is very broad, not well-understood, and operationalized differently across projects. Few studies have explored how clinical researchers define and operationalize engagement in DMHI research. Objective: This study investigated how clinical researchers operationalize user engagement with DMHIs in academic settings. Investigating operationalization can help identify gaps and inform strategies to better operationalize engagement in DMHIs according to intervention goals. Methods: We conducted exploratory qualitative semistructured interviews via Zoom (Zoom Communications, Inc; May 2023 to February 2024) with 12 clinical mental health researchers who had developed DMHIs using human-centered design methods. We recruited participants via purposive and snowball sampling. The interviews focused on understanding what participants considered engagement, how they measured it, the strategies they used to support or increase engagement, and the barriers they faced. We inductively coded the transcripts and conducted thematic analysis, iterating on codes and themes collaboratively. Results: We identified 3 dimensions of engagement for DMHIs: digital mental health components (ie, intervention, technology, and human support), levels of engagement (micro and macro), and visibility of the engagement (visible and invisible). We also described the challenges of designing DMHIs for engagement. Participants described components as overlapping; some viewed the technology and intervention as one, while others viewed the technology as distinct. Users should experience components as integrated. Within the levels of engagement, clinical researchers focused on designing for macroengagement but primarily measured microengagement through quantitative measures. Participants distinguished the visibility of engagement between what was capturable and measurable (visible engagement) and what could not be captured (invisible engagement). A major barrier to engagement was overcoming the invisibility of opportune moments for users to engage with the DMHI. Conclusions: Clinical mental health research focuses on end point clinical outcomes, while user engagement refers to the dynamic interaction with DMHIs in real-world settings, encompassing behavioral, cognitive, and affective user involvement. This tension highlights the need to operationalize engagement to better understand how users engage with DMHIs. A mixed method approach to capturing engagement would better align with clinical researchers' goals of understanding macroengagement. The dimensions of engagement (components, levels, and visibility) extend previous conceptualizations by providing support at the start of designing for the engagement of DMHIs. While capturing invisible aspects can still be challenging, awareness of the visibility of engagement can help researchers align their measurement and evaluation approaches with their macroengagement goals, instead of focusing on aspects of engagement that are readily visible.