The Role of Engagement with Mobile Health Interventions to Support Diabetes

The prevalence of type 2 diabetes (T2D) is growing significantly, yet fewer than half of American adults diagnosed with T2D are meeting target glycemic levels. Moreover, between 36% and 69% of adults with T2D have elevated diabetes distress. Recent advances have resulted in the proliferation of convenient and cost-effective technology-based interventions to support self-care in patients with T2D. In fact, the American Diabetes Association recommends multi-component technology plus coaching interventions to target self-care. Characterizing how individuals engage with technology-based interventions and how engagement is associated with outcomes is a critical next step. Family/friends Activation to Motivate Self-care (FAMS) is a multi-component self-care support intervention delivered by mobile phones that was recently studied in a randomized controlled trial. While FAMS improved self-efficacy, diabetes distress, global well-being, helpful involvement, and dietary behavior, patterns in engagement with the intervention’s various components have not been explored, nor has the role of engagement in mediating improvements observed in key target outcomes. Basic univariate descriptive statistics reveal meaningful variation in patient engagement (e.g., response to two-way messages and participation in coaching sessions), both across study participants and within participants longitudinally. The FAMS dataset presents an ideal opportunity to develop new methods in the space of engagement with multi-component interventions among adults with T2D. Specifically, our first aim is to study patterns in and predictors of engagement in multi-faceted interventions using innovative methods. Our second aim is to extend our previously developed modified instrumental variable approach to quantify the role of engagement in mediating the effects of FAMS relative to control. Understanding sources of heterogeneity in engagement and identifying its role in improving outcomes can help us personalize interventions to maximize treatment effects and fulfill the promise of digitally delivered interventions. Our investigation of the FAMS data will produce preliminary data necessary for an R01 submission in which we seek to further develop and implement methods to maximize engagement with multi-component interventions and identify optimal treatment strategies.