Attrition in Psychological mHealth Interventions for Young People: A Meta-Analysis

Abstract

AbstractPsychological interventions provided via mobile, wireless technological communication devices (mHealth) are a promising method of healthcare delivery for young people. However, high attrition is increasingly recognised as a challenge to accurately interpreting and implementing the findings of mHealth trials. The present meta-analysis consolidates this research and investigates potential methodological, sociodemographic, and intervention moderators of attrition. A systematic search of MEDLINE, PsycInfo, and Embase was conducted. Study reporting quality was evaluated (QualSyst tool), and attrition rates (proportions) were calculated and pooled, using both random- and mixed-effects models. The pooled attrition rate, sourced from 15 independent samples (Nparticipants = 1766), was 17% (CI [9.14, 30.13]). This increased to 26% (CI [15.20, 41.03]) when adjusting for publication bias. Attrition was significantly higher among application-based interventions (26%, CI [14.56, 41.86]) compared to those delivered via text or multimedia message (6%, CI [1.88, 16.98]). These data were, however, characterised by significant between-study variance. Attrition in mHealth trials with young people is common but may be mitigated by using message-based interventions. Taken together, the results can provide guidance in accounting for attrition across future mHealth research, clinical practice, public policy, and intervention design. However, sustained research focus on the effectiveness of different engagement strategies is needed to realise mHealth’s promise of equitable and efficient healthcare access for young people globally.