A systematic review and meta-analysis on digital mental health interventions in inpatient settings

Mental health disorders impose substantial personal and economic burdens, yet access to adequate treatment is limited due to various factors including provider shortages, negative attitudes towards treatment, and long waiting times. E-mental health (EMH) interventions, delivered through technology, offer potential solutions by overcoming geographical barriers, reducing waiting times, and increasing anonymity. Meta-analyses have established the effectiveness of EMH in outpatient settings for various disorders. However, research on EMH in inpatient settings is scarce, lacking a comprehensive meta-analytic synthesis. This study addresses this gap by conducting a meta-analysis to evaluate the efficacy of EMH interventions in inpatient settings, considering both blended interventions during inpatient stay and post-treatment aftercare.

Several meta-analyses demonstrate the effectiveness of EMH interventions in outpatient settings for conditions like anxiety, depression, eating disorders, PTSD, and work-related stress. Patient and practitioner acceptance of EMH is generally positive. In inpatient settings, timely and adequate intervention is crucial. EMH can supplement inpatient treatment by bridging waiting times, integrating with in-person interventions, or providing aftercare for relapse prevention. Existing research includes RCTs on EMH as add-ons to inpatient interventions and for aftercare, and a systematic review supported the efficacy of EMH aftercare, though limited by the small number of studies. However, to date, no meta-analyses have comprehensively examined EMH use in inpatient settings or specifically for aftercare.

This systematic review and meta-analysis followed PRISMA and Cochrane Handbook guidelines. Multiple databases (PubMed, ScienceGov, PsycInfo, CENTRAL, and others) were searched using keywords related to e-mental health, inpatient settings, and RCTs. Two literature searches were conducted (February and July 2024). Studies were included if they involved EMH interventions during inpatient treatment or aftercare, investigated mental health outcomes in inpatients, were RCTs, and provided sufficient data. Studies with high risk of bias (assessed using the Cochrane risk-of-bias tool for randomized trials, RoB 2) were excluded. Hedges' g was used to calculate effect sizes. Heterogeneity was anticipated and addressed using random-effects models. Publication bias was assessed using funnel plots and p-curve analyses. Post-hoc analyses examined the effects of EMH medium (e.g., web-based, app-based), type of control group, and specifically the effect on anxiety symptoms.

The meta-analysis included 26 studies (n=6112). Five studies used blended EMH during inpatient stay, while 21 used post-inpatient aftercare. The most common patient groups were eating disorders, mood disorders, and transdiagnostic groups. Various control groups were used (passive, active, active with added EMH). Funnel plot and p-curve analyses showed no evidence of publication bias. The overall effect size of EMH interventions was small but significant (g=0.3, 95% CI [0.2, 0.39]). Focusing only on clinically relevant outcomes increased the effect size (g=0.36). Blended interventions showed a larger effect (g=0.42) than aftercare (g=0.29). Significant effects were found for eating disorders, mood disorders, psychotic disorders, return-to-work interventions, and transdiagnostic groups. No significant effects were found for anxiety, somatic comorbidity, or substance abuse, though limitations in the number of studies investigating these conditions were noted. Post-hoc analyses revealed significant effects for web-based tools and multimedia interventions. EMH interventions were effective compared to both passive and active control groups. The effect size did not significantly decrease over the follow-up period (up to 24 months).

This meta-analysis provides the first evidence of the efficacy of EMH interventions in inpatient settings, showing a small but consistent positive effect across various disorders and intervention types. The effectiveness of EMH in inpatient aftercare is particularly noteworthy, as it addresses a critical need for relapse prevention in this high-risk population. The findings align with previous research demonstrating the efficacy of EMH in outpatient settings, suggesting that the benefits extend to inpatient contexts. The significant positive findings for psychotic disorders are unexpected given potential concerns, and warrant further investigation, particularly examining differences in EMH tool types.

EMH interventions are a valuable tool in inpatient mental health treatment and aftercare, particularly for mood, psychotic, and eating disorders. EMH can complement in-person treatment and offer support when in-person care is limited. Future research should focus on specific disorders, tool types, and larger-scale RCTs to confirm these findings and refine EMH implementation strategies. Investigations into engagement and adherence, and further research on under-represented conditions and populations are needed.

The meta-analysis is limited by the small number of studies, particularly for subgroup analyses. Many studies had some concerns regarding risk of bias. The limited number of studies on specific conditions (anxiety, substance abuse) prevents definitive conclusions. Geographical representation was predominantly Western/Northern Europe, limiting generalizability. Furthermore, the focus on symptom-focused measures limits insights into other relevant outcomes like remission rates.