Background: Digital suicide prevention interventions have previously been shown to be effective, however the rapid development of the field means a contemporary review of the evidence is warranted to help understand where interventions may work best. Additionally, recent years have seen a greater focus on lived experience participation in the development of interventions, without an assessment of the effectiveness of co-designed platforms on suicide ideation.
Methods: A meta-analysis following the PRISMA guidelines was conducted (PROSPERO CRD42022314246). PubMed/Medline, PsycINFO and Cochrane Central were searched for randomised controlled trials. Interventions were categorised according to their delivery setting, and as direct (directly targeting suicidality) or indirect (targeting depression), and effects on suicidal ideation and behaviours (plans, self-harm, attempts, suicide death) were extracted. In a post-hoc analysis, information about co-design, participatory design, or similar methods was extracted from manuscripts. A sub-group analysis was conducted comparing interventions which described such co-design approaches vs those which did not.
Results: Forty-six papers reporting 48 unique trials were included. The majority of studies examined direct interventions (n=27, 56.3%), and most were delivered in community settings (n=31, 61.3%). Overall, small and significant effects were observed for suicidal ideation (g: -0.15, p<0.001) but not suicidal behaviours (g: -0.17, p=0.794) at post-intervention. Small effect sizes, with overlapping confidence intervals, were found for clinical (g: -0.35, p=0.006), community (g: -0.10, p=0.037), and education settings (g: -0.20, p=0.283), although none of their pairwise comparisons reached significance. Small significant effects were observed for direct (g: -0.16, p=0.029) and indirect (g: -0.15, p=0.002) interventions, but again their pairwise comparison did not reach significance (z: 0.084, p=0.933). Interventions with co-design or similar approaches (n=10) experienced slightly larger effect sizes for suicidal ideation relative to interventions that did not involve co-design elements (g: -0.18 vs -0.15), although this difference was not significant (z: -0.295, p=0.767).
Conclusions: The results show that overall digital interventions for suicide prevention are effective, and that design features (such as treatment modality or delivery mode) may account for less variance in effectiveness than previously thought. Furthermore, while there is evidence for a greater effect on suicidal ideation when interventions involve some level of lived experience, this is non-significant when compared to interventions without a co-design element. We argue the quality of lived experience participation is highly variable, and a greater number of reliable studies with detailed descriptions of co-design processes is required in order to accurately determine its effectiveness.