Abstract
<title>Abstract</title> <p>Background Microbiota-directed therapies, including fecal microbiota transplantation (FMT), probiotics, and synbiotics, have emerged as candidate treatments for inflammatory bowel disease (IBD), yet no synthesis has simultaneously ranked these interventions across both efficacy and safety outcomes within a unified network. Methods We conducted a PRISMA-NMA–guided systematic review and network meta-analysis of randomized controlled trials evaluating microbiota-modifying therapies in adults with IBD. Bayesian random-effects NMA (binomial likelihood, logit link) and frequentist sensitivity analyses were performed for four outcomes: clinical remission, endoscopic remission, adverse events (AEs), and serious adverse events (SAEs). Local inconsistency was assessed by node-splitting; treatment hierarchies by SUCRA. Results Forty RCTs (2,722 patients) were included. For clinical remission, all active interventions except prebiotics were significantly superior to control: antibiotics + FMT (OR 9.53, 95% CrI 0.93–104.2), FMT (OR 2.81, 95% CrI 1.41–5.80), synbiotics (OR 2.61, 95% CrI 1.01–6.91), and probiotics (OR 2.29, 95% CrI 1.39–3.99). For endoscopic remission, only FMT reached significance in both models (OR 2.79, 95% CrI 1.45–5.38). No intervention significantly increased odds of AEs or SAEs. Node-splitting identified inconsistency in FMT-containing loops for clinical remission (p < 0.05). Heterogeneity was moderate for clinical remission (τ = 0.857) and low for SAEs (τ = 0.324). Conclusion FMT demonstrated the most consistent and robust efficacy across both remission outcomes, with no significant safety signal. Probiotics and synbiotics offered modest but significant clinical remission benefit. These findings support FMT as the preferred microbiota-directed strategy in IBD, pending resolution of network inconsistency through future head-to-head trials.</p>