JMIR Publications

ABSTRACT

Background:

Digital mental health interventions (DMHIs) are increasingly proposed to address rising mental health problems among university students, yet their effectiveness is persistently constrained by low and heterogeneous engagement. Conventional metrics such as initiation and module completion rates fail to capture how students actually experience, appropriate, and integrate these tools into their daily lives, and may mask meaningful differences between delivery formats.

Objective:

This study aimed to characterize the lived experience and mechanisms of engagement with ETUCARE, an 8-week DMHI for university students in France, and to examine how online-only and hybrid delivery formats (combining digital modules with face-to-face workshops) shape distinct engagement pathways and configurations.

Methods:

A qualitative-led convergent mixed-methods design was used. Semi-structured interviews with students from both conditions (N=23) were analyzed using reflexive thematic analysis (Braun & Clarke) to document engagement mechanisms, facilitators, and barriers. To corroborate and extend these findings, latent profile analysis (LPA) was conducted on a subsample of participants with detailed engagement data (N=97), integrating behavioral indicators derived from platform logs and experiential indicators (perceived quality, enjoyment, usefulness). Associations between delivery format and profile membership were examined using the R3STEP procedure, and ANOVAs compared conditions on composite engagement scores. Findings were integrated using a joint display.

Results:

Qualitative analysis identified six themes characterising how students engaged with ETUCARE. Online engagement was driven by self-regulated autonomy, flexible and selective use of a modular toolbox, and intermittent catch-up rhythms. Hybrid engagement was socially scaffolded, supported by psychological safety, group bonding, and facilitator guidance, but frequently burdened by the reframing of digital tasks as obligatory homework, workload underestimation, and compounded usability frictions. LPA corroborated these experiential patterns by identifying four distinct engagement profiles: engaged and satisfied (53.6%), engaged but dissatisfied (16.5%), low engagement (16.5%), and minimal engagement (13.4%). Hybrid participants had significantly higher odds of belonging to the low engagement profile rather than the engaged and satisfied profile (OR=3.51, 95% CI 1.07–11.52, P=.038), and showed significantly lower behavioral (ω²=.067) and experiential (ω²=.064) engagement than online-only participants.

Conclusions:

Engagement with ETUCARE was multidimensional and heterogeneous, shaped by distinct mechanisms across delivery formats. Hybrid delivery did not automatically enhance digital engagement and, under specific conditions, constrained it by transforming autonomous self-care into perceived obligation. These findings challenge prevailing assumptions about blended DMHIs and support mechanism-informed, qualitatively grounded approaches to measuring and optimizing engagement in youth digital mental health programs. Clinical Trial: Not applicable. The study was approved by the Ethics Committee of Nantes University (IRB No. IORG0011023, Ref. 06042023-2).