Expert Assessment of ChatGPT-4 Turbo-Generated Clinical Responses in Aneurysmal Subarachnoid Hemorrhage: A Systematic Quality Evaluation

This dataset contains the statistical analysis code and expert evaluation data for assessing the performance of ChatGPT-4 Turbo in generating responses to clinical questions on aneurysmal subarachnoid hemorrhage (aSAH). The study involved three cerebrovascular specialists who independently rated the AI-generated responses for accuracy, guideline adherence, and clinical relevance. The data includes descriptive statistics, inter-rater reliability measures (Fleiss’ kappa and Cronbach’s alpha), readability metrics (Flesch Reading Ease and Flesch-Kincaid Grade Level), as well as correlation and regression analyses. The results highlight the performance of ChatGPT-4 Turbo in addressing aSAH-related clinical questions, with insights into the linguistic complexity and guideline adherence of the generated responses. This dataset and code are intended to support further research into the integration of large language models in clinical decision-making.