Abstract
Background
Despite promising scalability and accessibility, evidence on the efficacy of self-guided interventions for adult depression is inconclusive. This study investigated their effectiveness and acceptability, considering diverse delivery formats and support levels.
Methods
We systematically searched PubMed, PsycINFO, Embase, and Cochrane Library until 1st January 2024. Included were randomised controlled trials comparing self-guided interventions with a control condition for adult depression. Two independent researchers extracted data. Effect sizes were pooled using random-effects models, with post-intervention depressive severity compared with control conditions as the primary outcome. Study validity was evaluated using Cochrane Risk of Bias 2.0. This study was pre-registered with OSF (https://osf.io/rd43v).
Findings
We identified 92 studies (111 interventions vs. control comparisons) with 16,706 participants (mean age: 18.78–74.41 years). Compared to controls, self-guided interventions were moderately effective at post-assessment (g = 0.53, 95% CI: 0.45–0.61; I2 = 79.17%) and six to twelve months post-randomisation follow-up (g = 0.32, 95% CI: 0.16–0.48; I2 = 79.19%). Trials with initial human screening (g = 0.59) and interventions delivered in computer programs (g = 1.04) had the significantly largest effect sizes. No differences in treatment effects were observed across support levels, therapy types, commercial availability, or the presence of online discussion forums. Self-guided interventions were less acceptable than control conditions (RR = 0.92, p < 0.001). Most studies showed a moderate to high risk of bias (n = 80).
Interpretation
Existing trials on self-guided interventions are at high risk of bias, potentially overestimating treatment effects. Despite lower acceptability compared to controls, self-guided interventions are moderately effective in treating adult depression, regardless of support levels and online discussion features.
Despite promising scalability and accessibility, evidence on the efficacy of self-guided interventions for adult depression is inconclusive. This study investigated their effectiveness and acceptability, considering diverse delivery formats and support levels.
Methods
We systematically searched PubMed, PsycINFO, Embase, and Cochrane Library until 1st January 2024. Included were randomised controlled trials comparing self-guided interventions with a control condition for adult depression. Two independent researchers extracted data. Effect sizes were pooled using random-effects models, with post-intervention depressive severity compared with control conditions as the primary outcome. Study validity was evaluated using Cochrane Risk of Bias 2.0. This study was pre-registered with OSF (https://osf.io/rd43v).
Findings
We identified 92 studies (111 interventions vs. control comparisons) with 16,706 participants (mean age: 18.78–74.41 years). Compared to controls, self-guided interventions were moderately effective at post-assessment (g = 0.53, 95% CI: 0.45–0.61; I2 = 79.17%) and six to twelve months post-randomisation follow-up (g = 0.32, 95% CI: 0.16–0.48; I2 = 79.19%). Trials with initial human screening (g = 0.59) and interventions delivered in computer programs (g = 1.04) had the significantly largest effect sizes. No differences in treatment effects were observed across support levels, therapy types, commercial availability, or the presence of online discussion forums. Self-guided interventions were less acceptable than control conditions (RR = 0.92, p < 0.001). Most studies showed a moderate to high risk of bias (n = 80).
Interpretation
Existing trials on self-guided interventions are at high risk of bias, potentially overestimating treatment effects. Despite lower acceptability compared to controls, self-guided interventions are moderately effective in treating adult depression, regardless of support levels and online discussion features.
| Original language | English |
|---|---|
| Article number | 105208 |
| Pages (from-to) | 1-10 |
| Number of pages | 10 |
| Journal | Ebiomedicine |
| Volume | 105 |
| Early online date | 14 Jun 2024 |
| DOIs | |
| Publication status | Published - Jul 2024 |