Author(s)
Andrew Cunningham, Medical Student
Hayley Behm, Medical Student
Shreyas Pyati, Medical Student
Aurelia S. Monk, BS, Medical Student
Charles S. Ebert Jr., MD, FARS
Brian Thorp, MD, FARS
Cristine Klatt-Cromwell, MD
Brent Senior, MD, FARS
Adam Kimple, MD, PhD, FARS
Affiliation(s)
Brody School of Medicine at East Carolina University;
Abstract:
Introduction: ChatGPT, a state-of-the-art language model, has the potential to be an easily accessible educational resource for patients. However, there is concern about the accuracy and safety of its medical information. We aimed to assess the utility of ChatGPT in patient education as compared to physician responses regarding rhinology topics.
Methods: ChatGPT was presented with questions related to common sinonasal complaints. Questions ranged from reasons for nasal congestion, treatment of chronic rhinosinusitis, and expectations or risks associated with sinus surgery. ChatGPT’s responses were distributed to a panel of experienced rhinologists, who evaluated the quality of responses on a Likert scale as well as identified errors, missing information, and aberrant medical advice, if applicable. A systematic literature review of artificial intelligence applications in rhinology was also conducted and summarized.
Results: ChatGPT's responses varied in quality. While the responses were generally accurate, there were several minor errors and occasionally suboptimal or incorrect advice. ChatGPT demonstrated potential in addressing straightforward questions but struggled with more nuanced and complicated questions.
Conclusion: ChatGPT is a potentially valuable resource for educating patients and providing quick answers to common inquiries. However, its limitations become evident when addressing intricate, multifaceted issues or complex medical decision-making. The integration of artificial intelligence tools in rhinology has the potential to enhance patient care, but careful consideration of their limitations and the need for expert oversight is essential in their implementation.