Abstract: Objective To explore the application of ChatGPT-assisted case-based learning (CBL) teaching in undergraduate clerkship teaching for undergraduate medical students. Methods Forty-nine medical students of grade 2020 from Peking University, rotating in the Anesthesiology Department of Peking University Third Hospital for clerkship, were enrolled in the study. They were randomized into an experimental group and a control group using the random number table method. The experimental group of 23 students received ChatGPT-assisted CBL teaching, while the control group of 26 students received traditional CBL training. After the course, teaching effectiveness was evaluated through knowledge test, skill assessment, and questionnaire survey results. Independent sample t-test and chi-square test were used for comparison between two groups. Results The knowledge test scores of the experimental group and the coutrol group of students were (78.30±8.49) points and (76.77±7.19) points, and the skill assessment scores were (92.87±3.95) points and (91.69±3.78) points, respectively (both P>0.05). The survey indicated that 13(50.0%) of the control group students hoped to incorporate ChatGPT into their CBL learning preparation; 14(60.9%) students in the experimental group believed that using ChatGPT in CBL teaching was more advantageous than other methods of searching for information; 13(56.5%) students believed that it could be promoted to other courses and would actively use it in the future. However, the accuracy of ChatGPT was lacking, especially for specific clinical problems, and it needed to be used discretionally. Conclusions The use of ChatGPT in CBL teaching can shorten preparation time and improve learning efficiency, but its accuracy is limited and requires instructor supervision.

Key words: Anesthesia, Artificial intelligence, ChatGPT, Case-based learning, Teaching method