MetaMind: A Multi-Agent Transformer-Driven Framework for Automated Network Meta-Analyses

Abstract Background Network meta-analysis (NMA) enables simultaneous comparison of multiple interventions by integrating direct and indirect evidence from randomized controlled trials and observational studies, but traditional workflows require extensive manual effort for study identification, data extraction, and statistical modelling, leading to slow update cycles and operational bottlenecks. Objective To develop and validate MetaMind , an end-to-end, transformer-driven framework that automates NMA processes—including study retrieval, structured data extraction, and meta-analysis execution—while minimizing human input. Methods MetaMind integrates Promptriever, a fine-tuned retrieval model, to semantically retrieve high-impact clinical trials from PubMed; a multi-agent LLM architecture--Mixture of Agents (MoA)-- pipeline to extract PICO-structured (Population, Intervention, Comparison, Outcome) endpoints; and GPT-4o–generated Python and R scripts to perform Bayesian random-effects NMA and other NMA designs within a unified workflow. Validation was conducted by comparing MetaMind’s outputs against manually performed NMAs in ulcerative colitis (UC) and Crohn’s disease (CD). Results Promptriever outperformed baseline SentenceTransformer with higher similarity scores (0.7403 vs. 0.7049 for UC; 0.7142 vs. 0.7049 for CD) and narrower relevance ranges. Promptriever performance achieved 82.1% recall, 91.1 % precision and an F 1 score of 86.4 % when compared to a previously published NMA. MetaMind achieved 100% accuracy in PICO element extraction and produced comparative effect estimates and credible intervals closely matching manual analyses. Conclusions MetaMind significantly reduces time and labour—delivering a complete NMA in under one week versus several months manually—while maintaining methodological rigor and scalability across therapeutic areas, representing a major advancement in AI-driven evidence synthesis