A systematic review and meta-analysis of transdiagnostic cognitive behavioural therapies for emotional disorders

Emotional disorders are highly prevalent and commonly co-occur. Shared cognitive, neuropsychological, and genetic factors underpin these disorders, supporting the rationale for transdiagnostic treatments. TD-CBT, an umbrella term for various approaches, addresses these shared mechanisms. Unified TD-CBT delivers the same treatment to patients regardless of specific diagnosis, targeting commonalities. This study addresses the existing gap in comprehensive meta-analyses by examining the efficacy of unified TD-CBT across individual, group, and internet-based settings, including both clinician-guided and unguided internet-based interventions. The study aims to provide an updated assessment of short and long-term TD-CBT efficacy compared to waitlist, treatment-as-usual (TAU), disorder-specific CBT (DS-CBT), and other active interventions, focusing on adult patients with emotional disorders and their anxiety and depression symptoms.

Prior reviews and meta-analyses of TD treatments have varied in their scope, considering different settings (group, individual, internet-based), populations (anxiety, anxiety and depression, or broader emotional disorders), and definitions of transdiagnostic approaches (unified, tailored, or specific protocols). Some focused on face-to-face anxiety treatments, while others compared TD treatments to disorder-specific treatments or exclusively examined internet-based interventions. A recent meta-analysis aggregated findings on transdiagnostic treatments for depression but didn't focus exclusively on CBT or include anxiety or other emotional disorders. Previous comprehensive meta-analyses on TD-CBT for emotional disorders are limited, either due to outdated search dates (2013) or inclusion of non-RCTs and exclusion of self-guided internet-based treatments. This study bridges this gap by providing a comprehensive, updated review encompassing various TD-CBT protocols and treatment settings.

This systematic review and meta-analysis followed the Cochrane Handbook and PRISMA guidelines. The PROSPERO protocol (CRD42019141512) was registered on September 27, 2019. PubMed, MEDLINE, PsycINFO, Google Scholar, medRxiv, and OSF Preprints were searched up to June 16, 2023. The search string combined terms for 'transdiagnostic', 'CBT', 'emotional disorder', and 'RCT'. Studies published between January 2000 and June 2023 were considered. Inclusion criteria involved adult populations with at least one clinician-established diagnosis of an emotional disorder (including anxiety, depressive, and adjustment disorders), unified TD-CBT interventions (excluding tailored treatments) based on CBT principles, and RCT designs with continuous self-report measures of anxiety and/or depression at pre- and post-treatment and follow-up. Two reviewers independently screened studies, extracted data, and assessed risk of bias using the revised Cochrane risk-of-bias tool (RoB 2.0). Controlled effect sizes (Hedges' g) were calculated using random-effects models, addressing heterogeneity through subgroup analyses (individual, group, internet-based) and outlier removal. Uncontrolled effect sizes (dSMC) were also calculated from pre- to post-treatment and follow-up assessments. Publication bias was assessed using funnel plots, rank correlations, Egger's test, and the Trim and Fill procedure.

The search identified 56 eligible RCTs (6,916 participants). After exclusions (data unavailability or lack of self-report measures), 53 studies (6,705 participants) were included in the meta-analysis. TD-CBT demonstrated significantly larger effect sizes for depression (g = 0.74, 95% CI = 0.57–0.92, P < 0.001) and anxiety (g = 0.77, 95% CI = 0.56–0.97, P < 0.001) compared to controls at post-treatment. TD-CBT was superior to waitlist and TAU across all treatment formats. Compared to DS-CBT, TD-CBT showed comparable efficacy for depression but not anxiety. Against other active treatments, TD-CBT showed greater efficacy for depression but not anxiety. The superiority of TD-CBT over controls was maintained at 3, 6, and 12-month follow-ups but not at 24 months. High heterogeneity was observed across studies, which decreased when stratifying by treatment format and removing outliers. Publication bias was detected. Bayesian analyses corroborated the frequentist findings. Comparable effects were seen across individual, group, and internet-based settings.

The findings strongly support the efficacy of TD-CBT for reducing symptoms of anxiety and depression in adults, across various formats. The large effect sizes, especially when compared to waitlist controls, are consistent with previous research. The comparable efficacy of TD-CBT to DS-CBT, particularly for depression, challenges previous mixed findings and highlights its potential as a viable alternative. The sustained effectiveness observed at 3, 6, and 12 months, combined with evidence of comparable long-term outcomes (up to 24 months for some studies), further underscores TD-CBT's potential. The comparable effects across treatment settings (individual, group, and internet-based) suggest its flexibility and potential for widespread implementation, particularly in resource-constrained environments. Internet-based TD-CBT shows promise for reaching underserved populations.

This meta-analysis provides robust evidence for the efficacy of TD-CBT for emotional disorders, offering a potentially cost-effective and broadly applicable treatment across different settings. Future research should focus on individual participant data meta-analyses to explore potential moderators (treatment dose, patient characteristics, protocol variations), investigating clinical relevance of symptom improvements and exploring outcomes beyond anxiety and depression. Further investigation of contraindications and potential for personalized TD-CBT is also warranted.

Limitations include the focus on self-reported outcomes, leading to potential bias in blinding. Heterogeneity across studies suggests a need for more research investigating factors such as specific protocols and treatment intensity. Underrepresentation of studies from South America and Africa limits the generalizability of findings. Publication bias was noted, and the lack of a priori specified analysis plans in many studies raises concerns about selective reporting. The inclusion criteria limited the analysis to unified TD-CBT, excluding tailored interventions.