Abstract
Abstract Background Selecting an appropriate treatment strategy for inflammatory bowel disease (IBD) is an important task that can significantly alter the clinical course of the disease. Advanced artificial intelligence language models, such as ChatGPT, represent a potential future option that can assist practitioners. ChatGPT can help search the internet for medical facts by summarizing information from medical sources and providing concise explanations, simplifying the process of accessing and understanding medical information. Methods A ChatGPT-guided analysis of 89 IBD patients treated at the Clinical Department of Gastroenterology and Hepatology at University Medical Center "Zvezdara", Belgrade was conducted, considering further treatment with some biological drugs. A treatment modality proposal was obtained through communication with the interactive software ChatGPT (OpenAI, CA USA) and providing complete clinical and laboratory data. The alignment of decisions was analyzed, along with an assessment of treatment success in the group with matching and differing choices. Results In 72 patients (81%), there was concordance in the choice of therapy between ChatGPT and the medical consultation, while in 17 patients, ideas about the treatment approach differed. The most common difference in the choice of therapeutic modality related to the software's selection of ustekinumab compared to our choice of vedolizumab in patients with Crohn's disease who were not biologically naive, affecting seven patients with these characteristics. In 4 patients with severe ulcerative colitis, ChatGPT recommended using cyclosporine in combination with vedolizumab, contrary to the medical consultation's decision to use Anti-TNFα therapy. The group of patients in whom the artificial intelligence advice differed from the medical consultation was in a clinically more severe stage of the disease (CRP 10.4 vs. 9.1, p=0.0293), with worse endoscopic findings (proportion of severe endoscopic disease SES CD≥33/Mayo score>2, 53% vs. 24%, p=0.019). Taking into account the baseline differences in these patient groups, multiple analyses of covariance did not reveal a significant difference in the ultimate treatment outcome - proportions of endoscopic (p=0.773) and clinical remission (p=0.296) between patient groups. Conclusion The integration of advanced artificial intelligence language models may represent a future tactic with the potential to assist practitioners in making decisions about the treatment modality for IBD patients, but clinical experience should be considered.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
