Smartphone-Delivered Attentional Bias Modification Training for Mental Health: Systematic Review and Meta-Analysis (Preprint)

BACKGROUND Smartphone-delivered attentional bias modification training (ABMT) intervention has gained popularity as a remote solution for alleviating symptoms of mental health problems. However, the existing literature presents mixed results indicating both significant and insignificant effects of smartphone-delivered interventions. OBJECTIVE This systematic review and meta-analysis aims to assess the impact of smartphone-delivered ABMT on attentional bias and symptoms of mental health problems. Specifically, we examined different design approaches and methods of administration, focusing on common mental health issues, such as anxiety and depression, and design elements, including gamification and stimulus types. METHODS Our search spanned from 2014 to 2023 and encompassed 4 major databases: MEDLINE, PsycINFO, PubMed, and Scopus. Study selection, data extraction, and critical appraisal were performed independently by 3 authors using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. When necessary, we pooled the standardized mean difference with a 95% CI. In addition, we conducted sensitivity, subgroup, and meta-regression analyses to explore moderator variables of active and placebo ABMT interventions on reducing symptoms of mental health problems and attentional bias. RESULTS Our review included 12 papers, involving a total of 24,503 participants, and we were able to conduct a meta-analysis on 20 different study samples from 11 papers. Active ABMT exhibited an effect size (Hedges g) of –0.18 (P=.03) in reducing symptoms of mental health problems, while the overall effect remained significant. Similarly, placebo ABMT showed an effect size of –0.38 (P=.008) in reducing symptoms of mental health problems. In addition, active ABMT (Hedges g –0.17; P=.004) had significant effects on reducing attentional bias, while placebo ABMT did not significantly alter attentional bias (Hedges g –0.04; P=.66). CONCLUSIONS Our understanding of smartphone-delivered ABMT’s potential highlights the value of both active and placebo interventions in mental health care. The insights from the moderator analysis also showed that tailoring smartphone-delivered ABMT interventions to specific threat stimuli and considering exposure duration are crucial for optimizing their efficacy. This research underscores the need for personalized approaches in ABMT to effectively reduce attentional bias and symptoms of mental health problems. CLINICALTRIAL PROSPERO CRD42023460749; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=460749