| Popis: |
BACKGROUND Recent years have highlighted the increasing need to promote mental well-being in the general population. This has led to a rapidly growing market of fully automated digital mental well-being tools. Although many individuals have started using these tools in their daily lives, evidence on the effectiveness of digital mental well-being tools is currently lacking. OBJECTIVE The objective of the current study was therefore to review the currently available evidence on the effectiveness of fully automated digital interventions to promote mental well-being in the general population. METHODS Following preregistration of the systematic review protocol on PROSPERO (registration: CRD42022310702), searches were carried out in: Medline, Web of Science, Cochrane, PsychINFO, PsychEXTRA, Scopus and ACM Digital (initial searches in February 2022; updated in October 2022). Studies were included if they contained a general population sample and a fully automated digital intervention that exclusively employed mental well-being promotion activities. Two reviewers, blinded to each other’s decisions, conducted data selection, extraction and quality assessment of the included studies. A narrative synthesis and a random-effects model of Per Protocol (PP) data were adopted. RESULTS A total of 7,243 participants in 19 studies were included. These studies contained 24 fully automated digital mental well-being interventions of which 15 were included in the meta-analysis. Compared with no intervention, there was a significant small effect of fully automated digital mental well-being interventions on mental well-being in the general population (SMD = 0.19, 95% CI ranging from 0.04 to 0.33). Specifically, mindfulness, acceptance & commitment, and compassion-based interventions significantly promoted mental well-being in the general population; evidence was lacking on positive psychology and Cognitive Behavioural Therapy-based interventions; and contraindications were found for integrative approaches. Overall, there was substantial heterogeneity which could partially be explained by the intervention duration, comparator and study outcome. Risk of bias was high and confidence in quality of the evidence very low (GRADE), primarily due to the high rates of dropout (averaging 37%, ranging from 0-85%) and suboptimal intervention adherence (averaging 57%). CONCLUSIONS In conclusion, this study provides a novel contribution to knowledge regarding the effectiveness, as well as the strengths and weaknesses of fully automated digital mental well-being interventions in the general population. Future research and practice should take these findings into account when developing fully automated digital mental well-being tools. Additionally, research should aim to investigate further strategies to improve adherence and reduce dropout in fully automated digital mental well-being interventions and aim to understand when and for whom these interventions are particularly beneficial. CLINICALTRIAL PROSPERO registration: CRD42022310702 |
