Evaluating AI-Generated Responses for Patient Counseling and Treatment Decision-Making in Melasma: A Cross-Sectional...

Background Melasma poses challenges in treatment decisions and patient counseling due to its chronic nature and variable response to therapies. Evaluating AI-generated responses from ChatGPT (OpenAI, Inc., San Francisco, CA, USA) and ChatSonic (Writesonic, Inc., San Francisco, CA, USA) could provide insights into their potential role in supporting dermatologists with patient counseling and treatment decision-making for melasma. Aim This study aims to compare responses generated by ChatGPT and ChatSonic for patient counseling and treatment decisions in melasma, as evaluated by senior dermatologists for satisfaction, quality, and reliability. Methodology This cross-sectional observational study, conducted virtually in January 2023, involved asking five questions each about patient counseling and treatment decision-making in melasma to ChatGPT and ChatSonic. Responses were compiled in a Microsoft Excel assessment sheet (Microsoft Corporation, Redmond, WA, USA) and presented in a blinded format to two senior dermatologists (Evaluators A and B) for evaluation of quality (Global Quality Score), reliability (modified DISCERN scale), and satisfaction (Likert scale). Results For patient counseling in melasma, ChatGPT responses were rated higher in satisfaction by Evaluator A, with no significant differences in quality or reliability between the two AI chatbots. For treatment decisions, Evaluator B significantly favored ChatGPT over ChatSonic in both satisfaction and quality. Conclusions AI chatbots such as ChatGPT and ChatSonic can effectively support dermatologists in patient counseling and treatment decision-making for melasma, with ChatGPT demonstrating superior satisfaction and quality.