Evaluating ChatGPT's Ability to Address Frequently Asked Questions in Gender-Affirmation Surgery.

Abstract

ChatGPT has significantly influenced healthcare, yet its impact on patient education regarding gender-affirmation surgery (GAS) remains underexplored. This study aimed to evaluate ChatGPT's utility in providing medical information to patients seeking GAS. In the first part of the study, we collected questions from the "Ask a Surgeon" forum hosted by the American Society of Plastic Surgery and compared responses from verified physicians on the forum to those generated by ChatGPT. We found that ChatGPT's responses were significantly more complex across five readability metrics but had significantly higher DISCERN and PEMAT scores compared to physician responses, indicating superior reliability, quality, and understandability. In the second part of our study, ChatGPT was queried using ten frequently asked questions to simulate a patient's experience seeking treatment information. ChatGPT's responses were generally detailed and on-topic, emphasized the importance of consulting a healthcare provider, and highlighted the psychological and emotional factors associated with GAS. Overall, ChatGPT showed promise as an effective tool for patient education in GAS. It provides clear, private information, correctly emphasizes the psychosocial needs of this patient population, and consistently advises consultation with healthcare professionals. However, its high reading level and lack of transparent references raise concerns about its implementation.