Digital Interventions for Symptoms of Borderline Personality Disorder: Systematic Review and Meta-Analysis

Background: Borderline personality disorder (BPD) is a mental health condition with insufficient care availability worldwide. Digital mental health interventions could reduce this treatment gap. Persuasive system design (PSD) is a conceptual framework outlining elements of digital interventions that support behavior change. Objective: This systematic review aims to characterize digital interventions targeting BPD symptoms, assess treatment efficacy, and identify its association with intervention features, including PSD elements. Methods: A systematic review of automated digital interventions targeting symptoms of BPD was conducted. Eligible studies recruited participants aged >= 18 years, based on a diagnosis of BPD or one of its common comorbidities, or as healthy volunteers. OVID Embase, OVID MEDLINE, OVID PsycINFO, and the Cochrane Central Register for Controlled Trials were searched on July 19, 2022, and February 28, 2023. Intervention characteristics were tabulated. A meta-analysis of randomized controlled trials (RCTs) determined treatment effects separately for each core symptom of BPD using Hedges g. Associations between the treatment effect and intervention features, including PSD elements, were assessed by subgroup analysis (Cochran Q test). Risk of bias was assessed using the Cochrane Risk of Bias 2 tool for RCTs and the National Institutes of Health Quality Assessment Tool for pre-post studies. Results: A total of 40 (0.47%) publications out of 8520 met the inclusion criteria of this review, representing 6611 participants. Studies comprised examinations of 38 unique interventions, of which 32 (84%) were RCTs. Synthesis found that included interventions had the following transdiagnostic treatment targets: severity of BPD symptoms (4/38, 11%), suicidal ideation (17/38, 45%), paranoia (5/38, 13%), nonsuicidal self-injury (5/38, 13%), emotion regulation (4/38, 11%), and anger (3/38, 8%). Common therapeutic approaches were based on dialectical behavioral therapy (8/38, 21%), cognitive behavioral therapy (6/38, 16%), or both (5/38, 13%). Meta-analysis found significant effects of digital intervention for both symptoms of paranoia (Hedges g=-0.52, 95% CI -0.86 to -0.18; P=.01) and suicidal ideation (Hedges g=-0.13, 95% CI -0.25 to -0.01; P=.03) but not overall BPD symptom severity (Hedges g=-0.17, 95% CI -0.42 to 0.10; P=.72). Subgroup analysis of suicidal ideation interventions found that evidence-based treatments such as cognitive behavioral therapy and dialectical behavior therapy were significantly more effective than alternative modalities (Cochran Q=4.87; P=.03). The degree of human support was not associated with the treatment effect. Interventions targeting suicidal ideation that used reminders, offered self-monitoring, and encouraged users to rehearse behaviors were associated with a greater reduction in ideation severity. Conclusions: Evidence suggests that digital interventions may reduce the symptoms of suicidal ideation and paranoia and that the design of digital interventions may impact the efficacy of treatments targeting suicidal ideation. These results support the use of transdiagnostic digital interventions for paranoia and suicidal ideation.