Garcia Garcia Laura, Emile Sameh Hany, Linkeshwaran Lakshmi, Wignakumar Anjelli, Wexner Steven D
Servicio de Cirugía General y Digestiva. Complejo Hospitalario Universitario Materno Infantil de Gran Canaria, Las Palmas de Gran Canaria, Spain.
Ellen Leifer Shulman and Steven Shulman Digestive Disease Center, Cleveland Clinic Florida, Weston, FL; Colorectal Surgery Unit, General Surgery Department, Mansoura University Hospitals, Mansoura, Egypt. Electronic address: https://twitter.com/dr_samehhany81.
Surgery. 2025 Jul;183:109393. doi: 10.1016/j.surg.2025.109393. Epub 2025 May 9.
Artificial intelligence-based chatbots are becoming increasingly used in patient education, in the realm of colorectal diseases. Perhaps, not surprisingly, concerns about the appropriateness of chatbot answers have been raised by healthcare professionals. Numerous studies have explored the utility and accuracy of chatbots in providing information in several clinical disciplines. This review aimed to summarize the findings of published studies, highlighting the strengths and limitations of chatbots used in medical education for colorectal surgery.
We searched MEDLINE via PubMed and Scopus in February 2025 for original articles evaluating artificial intelligence-based chatbots in patient education related to colorectal surgery, categorizing them into 3 groups: colorectal cancer, inflammatory bowel diseases, and other colorectal conditions.
We identified 15 studies, 9 assessed chatbot utility in patient education in colorectal cancer, 4 assessed their utility in inflammatory bowel diseases, 1 involved benign anal conditions, and another involved intestinal stomas. Our findings indicated that chatbots, particularly ChatGPT, can improve patient education by providing accessible information on common questions. However, we also identified several limitations of the ability of chatbots to address complex medical issues which underscored that these tools may complement rather than replace professional medical guidance.
Chatbots may be useful for patient education related to simple and basic information, but not in complex and patient-specific settings. Future research should focus on refining chatbot algorithms to enhance the accuracy and depth of their responses, ensuring they effectively support patient education while maintaining the crucial role of healthcare providers.
基于人工智能的聊天机器人在结直肠疾病领域的患者教育中越来越常用。或许并不奇怪的是，医疗保健专业人员对聊天机器人答案的恰当性提出了担忧。众多研究探讨了聊天机器人在多个临床学科提供信息方面的实用性和准确性。本综述旨在总结已发表研究的结果，突出用于结直肠外科医学教育的聊天机器人的优势和局限性。
我们于2025年2月通过PubMed和Scopus检索MEDLINE，查找评估基于人工智能的聊天机器人在结直肠手术相关患者教育中的原始文章，并将其分为3组：结直肠癌、炎症性肠病和其他结直肠疾病。
我们确定了15项研究，9项评估了聊天机器人在结直肠癌患者教育中的实用性，4项评估了其在炎症性肠病中的实用性，1项涉及良性肛门疾病，另一项涉及肠造口。我们的研究结果表明，聊天机器人，尤其是ChatGPT，可以通过提供常见问题的易获取信息来改善患者教育。然而，我们也发现了聊天机器人处理复杂医疗问题能力的一些局限性，这凸显出这些工具可能是补充而非取代专业医疗指导。
聊天机器人可能对与简单和基本信息相关的患者教育有用，但在复杂和针对特定患者的情况下则不然。未来的研究应专注于完善聊天机器人算法，以提高其回答的准确性和深度，确保它们在维持医疗保健提供者关键作用的同时有效支持患者教育。
