VVC Large Language Model Chatbot Evaluation: Supplementary Materials and Dataset

This record contains the supplementary materials and research dataset supporting the study entitled “Safety, Accuracy, Empathy, Information Quality, and Readability of Six Large Language Model Chatbots for Vulvovaginal Candidiasis Advice.” The deposited package includes the completed CHART checklist, the reconstructed question-screening log, the evidence-based question–answer mapping table, the complete record of 372 chatbot-generated responses to 62 patient-oriented questions, the statistical supplement, a qualitative safety-risk analysis, readability and statistical analysis code written in R, original expert-rating data, supplementary tables, and Figures 1–4. The chatbot outputs are provided as research data and should not be interpreted as reference-standard medical advice. Some responses may contain inaccuracies, clinically important omissions, or potentially unsafe recommendations. No patient data, medical records, biological specimens, or identifiable personal information are included. Version: 1.1.0.