Comparison of Clinical Assessment and Large Language Models in Preoperative Risk Classification: A Retrospective Analysis of ChatGPT, DeepSeek, Gemini, and Claude in ASA Physical Status Classification

Can AI Outperform Clinicians on ASA Physical Status Classification? New Retrospective Trial Pits ChatGPT, DeepSeek, Gemini, and Claude Against Senior Anesthesiologists

A new retrospective trial from Marmara University will benchmark ChatGPT, DeepSeek, Gemini, and Claude against senior anesthesiologists in assigning ASA Physical Status classifications from de-identified clinical vignettes. With interrater variability in ASA-PS classification a long-standing clinical problem, the findings could have significant implications for perioperative risk assessment practice worldwide.