Artificial Intelligence and Systemic Isotretinoin: Patient Education on Ophthalmological Side Effects

Background/Purpose: Isotretinoin is frequently utilized in the treatment of skin disorders such as acne vulgaris and rosacea, with its associated side effects occurring relatively often. The most prevalent ophthalmological adverse effects include conjunctivitis, hordeolum, chalazion, blepharitis, and xerophthalmia. In light of the growing integration of artificial intelligence within healthcare, this study aims to evaluate the capability of ChatGPT to enhance patient education regarding these conditions. Materials and Methods: A questionnaire consisting of fifteen frequently asked questions concerning the ocular side effects of systemic isotretinoin was developed. Eight questions, selected by a dermatologist and a ophthalmologist, were posed to ChatGPT version 4.0. The responses were assessed using a four-point rating scale. Answers were independently categorized as "excellent," "satisfactory with minimal explanation required," "satisfactory with moderate explanation required," or "unsatisfactory." Results: ChatGPT provided accurate and informative responses to all eight questions presented. Six responses were rated as "excellent," while two were categorized as "satisfactory with minimal explanation required." Inter-rater reliability was assessed using Cohen's kappa analysis, resulting in a kappa value of 0.413 (95% confidence interval, 0.007 to 0.825). Conclusion: ChatGPT exhibited a significant capacity to effectively address patient inquiries related to the ophthalmological side effects of systemic isotretinoin. Following a review by dermatologists and ophthalmologists, it may serve as a valuable complementary tool for patient education.