Performance of ChatGPT as an AI-assisted decision support tool in medicine: a proof-of-concept study for interpreting symptoms and management of common cardiac conditions (AMSTELHEART-2)

Abstract

ABSTRACTBackgroundIt is thought that ChatGPT, an advanced language model developed by OpenAI, may in the future serve as an AI-assisted decision support tool in medicine.ObjectiveTo evaluate the accuracy of ChatGPT’s recommendations on medical questions related to common cardiac symptoms or conditions.MethodsWe tested ChatGPT’s ability to address medical questions in two ways. First, we assessed its accuracy in correctly answering cardiovascular trivia questions (n=50), based on quizzes for medical professionals. Second, we entered 20 clinical case vignettes on the ChatGPT platform and evaluated its accuracy compared to expert opinion and clinical course.ResultsWe found that ChatGPT correctly answered 74% of the trivia questions, with slight variation in accuracy in the domains coronary artery disease (80%), pulmonary and venous thrombotic embolism (80%), atrial fibrillation (70%), heart failure (80%) and cardiovascular risk management (60%). In the case vignettes, ChatGPT’s response matched in 90% of the cases with the actual advice given. In more complex cases, where physicians (general practitioners) asked other physicians (cardiologists) for assistance or decision support, ChatGPT was correct in 50% of cases, and often provided incomplete or inappropriate recommendations when compared with expert consultation.ConclusionsOur study suggests that ChatGPT has potential as an AI-assisted decision support tool in medicine, particularly for straightforward, low-complex medical questions, but further research is needed to fully evaluate its potential.