Clinical and demographic predictors of engagement in mobile health vs. clinic-based interventions for serious mental illness

Abstract

Evidence-based interventions for individuals with serious mental illness (SMI) depend on sustained engagement, but individuals with SMI often disengage from treatment prematurely. Mobile health (mHealth) interventions may alleviate barriers that are associated with in-person services. In the present study we evaluate whether demographic and clinical characteristics of patients with SMI (n = 163) are differentially associated with their engagement when randomized to mHealth vs. clinic-based care. With regard to diagnosis and demographics, participants randomized to mHealth engaged for more weeks across predictors; however, exploratory subgroup analyses suggested differences in engagement were highest among individuals with bipolar disorder (d = .97) and those who had previously used a smartphone (d = .72). With regard to symptom severity, differences in engagement between conditions were greatest amongst individuals with elevated depressive (d = .99) and psychotic (d = 1.01) symptoms as well as lower self-reported recovery (d = .85). Differences in engagement between mHealth and in-person care amongst those with highest self-rated recovery and lowest depressive and psychotic symptoms were small and non-significant. These findings suggest that mHealth may have an especially large advantage in maintaining engagement amongst individuals with the most severe clinical presentations. This suggests mHealth could provide added value in keeping individuals at elevated risk of poor outcomes connected to care. The main trial was registered in clinicaltrials.gov (NCT02421965; https://clinicaltrials.gov/ct2/show/NCT02421965).