Artificial intelligence in gastroenterology clinical practice: Scoping review of large language model applications.

Clinical Snapshot

Key Findings

• Effect Size: Not applicable — scoping review with thematic synthesis; no pooled effect size computed

• Primary Outcome: Mapping of LLM applications across gastroenterology subspecialties and clinical domains: 73 eligible studies identified from 2,895 screened. Six subspecialties represented; hepatology (20/73, 27.3%) and endoscopy (17/73, 23.2%) most studied. Six application domains identified; patient education and communication (38 studies) and decision support and clinical guidance (24 studies) most frequent.

• Nnt Or Sensitivity: Not applicable — no diagnostic accuracy meta-analysis or therapeutic NNT calculable from this scoping review design; individual study-level accuracy metrics are reported narratively but not pooled

• Confidence Interval: Not applicable — no quantitative meta-analysis performed

Clinical Application

Feasibility of LLM integration in gastroenterology remains unproven at the clinical deployment level. Most evidence derives from simulation studies. Real-world implementation requires validation of accuracy in live clinical environments, integration with electronic medical record systems, clinician training, and governance frameworks for AI-assisted decision-making. The review identifies growing but still nascent real-world data. In Australia, LLM deployment in clinical settings falls under the TGA's evolving Software as a Medical Device (SaMD) regulatory framework, with the TGA having released guidance on AI/ML-based medical devices. The RACGP and Gastroenterological Society of Australia (GESA) have not yet issued specific guidelines on LLM use in gastroenterology practice. PBS implications are indirect at this stage, though LLM-assisted documentation and triage could influence referral patterns and resource utilisation. Australian clinicians should note that most evaluated models (GPT-3.5, GPT-4) are US-developed and may not reflect Australian clinical guidelines, PBS formulary, or local epidemiological patterns (e.g., high rates of NAFLD/MASLD, colorectal cancer screening via the National Bowel Cancer Screening Program). Any clinical deployment would require local validation and ethics approval. Gastroenterologists, hepatologists, endoscopists, and gastroenterology nursing and allied health staff considering integration of LLM-based tools into clinical workflows for patient education, documentation support, or decision assistance. Also relevant to clinical informaticists and health service administrators evaluating AI procurement.

Abstract

BACKGROUND: Advances in artificial intelligence have brought renewed attention to tools that can work with the large amount of written information generated in clinical practice. Among these, large language models (LLMs) stand out for their ability to interpret and generate medical text in a flexible, context-aware way. Gastroenterology, like many specialties, produces a wide range of narrative and semi-structured data, and this has encouraged researchers to explore how LLMs might help clinicians manage everyday tasks. Recent studies have examined their potential contributions to patient education, communication between care teams and patients, decision support, and routine documentation, reflecting a growing interest in how these systems might fit into real-world clinical workflows. OBJECTIVE: This scoping review aimed to map current applications of LLMs in gastroenterology clinical practice, including subspecialty focus, study designs, model types, and reported outcomes. METHOD: Following PRISMA guidelines, a systematic search was conducted in PubMed, Scopus, and Web of Science for studies published between January 2022 and August 2025. Eligible studies included original research assessing LLM applications in gastroenterology clinical practice. Data were extracted on subspecialty, application domain, LLM type, data source, and outcomes. We employed thematic analysis to address our primary research question. RESULTS: 73 out of 2895 studies identified in the initial search met the inclusion criteria. Six subspecialties and six application domains emerged from our review. Hepatology (20/73 studies, 27.3%) and endoscopy (17/73 studies, 23.2%) were the most represented subspecialties. The most frequently investigated application domains were patient education and communication (38 studies) and decision support and clinical guidance (24 studies). Most studies were simulation-based or literature-based cases, although an increasing number have used real-world clinical data, particularly in recent years. The majority evaluated general-purpose models such as GPT-3.5 and GPT-4, with some incorporating retrieval augmentation or fine-tuning. Reported outcomes varied by application domain and included measures of accuracy, concordance, completeness, relevance, safety, reliability, usability, user satisfaction, efficiency, time savings, and educational value. Commonly described limitations included variable reliability, incomplete responses, and challenges in generalizing from simulated to clinical settings. CONCLUSIONS: Research on LLMs in gastroenterology has expanded across multiple subspecialties and application domains. Current evidence is primarily based on simulation studies, with limited but growing evaluation using real-world clinical data. Further work is needed to assess performance in prospective and applied clinical contexts.

References

1. 1.Yazarkan, Y., Sonmez, G., & Simsek, C. (2026). Artificial intelligence in gastroenterology clinical practice: Scoping review of large language model applications. International Journal of Medical Informatics. https://doi.org/10.1016/j.ijmedinf.2026.106413