Face-to-Face Versus Digital, Telephone-Delivered, and Self-Help Cognitive Behavioral Therapy for Irritable Bowel Syndrome: Systematic Review and Bayesian Indirect Treatment Comparison Meta-Analysis.

Abstract

Cognitive behavioral therapy (CBT) is recommended for irritable bowel syndrome (IBS). However, it remains unclear whether face-to-face CBT is as effective as digital, self-help, or telephone-delivered CBT for IBS. This study aimed to estimate the relative effects of face-to-face CBT compared with digital, telephone-delivered, and self-help CBT for IBS and to assess whether there are adequate effective sample sizes to support the findings. Ovid MEDLINE, Embase, and the Cochrane Library were searched up to September 27, 2025. Randomized controlled trials of face-to-face, digital, self-help, or telephone-delivered CBT for IBS in adults were included. The primary outcome was the IBS symptom severity scale. The secondary outcomes were IBS quality of life and abdominal pain intensity. A Bayesian random effects model was used for the meta-analysis. The effective and required sample sizes were calculated to estimate whether the sample sizes were adequate. The certainty of evidence was evaluated using the Confidence in Network Meta-Analysis Framework. The risk of bias of included studies was assessed using the Cochrane Collaboration's risk of bias tool (version 2). We analyzed 22 studies involving 3161 participants. The number of participants ranged between 28 and 558. The mean (SD) age of participants was 37.2 (10.6) years, and 78.6% (2485/3161) were women. These randomized controlled trials were published between 2003 and 2025. We found that face-to-face CBT had similar effects compared with digital CBT (mean difference [MD] -0.89, 95% credible interval [CrI] -20.78 to 18.73), self-help CBT (MD -1.73, 95% CrI -21.03 to 17.80), and telephone-delivered CBT (MD -0.76, 95% CrI -20.86 to 19.38) in improving IBS symptom severity scale scores. The comparison between face-to-face CBT and self-help CBT had sufficient effective sample sizes (375/140), whereas the effective sample sizes for comparisons with digital CBT (347/729) and telephone-delivered CBT (140/627) were insufficient. The certainty of evidence was moderate to low. Similarly, in improving quality of life and abdominal pain intensity, face-to-face CBT showed equal effect compared with digital and self-help CBT, with insufficient sample sizes and low to very low evidence certainty. This is the first Bayesian meta-analysis to incorporate effective and required sample size calculations for comparisons among CBT modalities in IBS. We analyzed continuous data of the outcomes. Meanwhile, we computed the effective and required sample sizes, thereby quantifying the informational adequacy of each comparison. Our Bayesian meta-analysis demonstrated significant potential for digital, self-help, and telephone-delivered CBT for patients with IBS, but the effective sample sizes of most comparisons were inadequate. Digital, self-help, and telephone-delivered CBT can serve as important options for managing IBS in clinical practice. Given high heterogeneity, high risk of bias, and inadequate effective sample sizes, more high-quality studies are warranted.