DIGITAL EXPERTS IN LYMPHEDEMA: ASSESSING THE QUALITY AND READABILITY OF RESPONSES FROM CHATGPT AND GEMINI

This study evaluated the quality and readability of responses provided by ChatGPT and Gemini to frequently asked questions related to lymphedema. Ten frequently asked questions about lymphedema were selected by expert therapists for submission to ChatGPT and Gemini. The initial responses from ChatGPT and Gemini were recorded without follow-up queries. Five independent experts (therapists) specialized in this field evaluated responses from ChatGPT and Gemini using a four-point rating scale. Readability levels were analyzed using the Flesch-Kincaid Grade Level through WordCalc software. Results indicated that except for questions 3 and 6, ChatGPT and Gemini provided similar responses, with significant differences observed in those two questions (p = 0.014). ChatGPT’s answers to questions 1, 2, 3, 4, 8, 9, and 10 were found to be more readable than Gemini’s, while Gemini’s answers to questions 5, 6, and 7 were more readable. Overall, approximately 70% of ChatGPT’s responses were found to be easier to read compared to Gemini. The results of this study indicate that both AI search engines offer comparable responses to research questions. However, when evaluating clarity and accessibility of the answers, ChatGPT was found to be more understandable and user-friendly than Gemini.