Researchers at Beth Israel Deaconess Medical Center (BIDMC) have conducted a study comparing a chatbot’s probabilistic reasoning to that of human clinicians. The findings suggest that AI could serve as valuable clinical decision support tools for physicians. The study was based on a survey of over 550 practitioners who performed probabilistic reasoning on five medical cases. The chatbot’s task was to estimate the likelihood of a diagnosis based on patient presentations and update its estimates when provided with test results. The chatbot demonstrated similar or better accuracy in making diagnoses compared to human practitioners, especially when test results were negative. The chatbot’s consistency in such scenarios may help reduce unnecessary interventions. Dr. Rodman emphasized the potential of chatbots to support highly skilled physicians in their clinical decision-making.
