Comparative study of large language model-assisted learning and traditional text-based learning in flap transplantation education
CHEN Hongyu
LI Zijun
PAN Xingyi
JIN Mengying
AN Yang
Abstract：Objective To directly compare the immediate effects of large language model-based conversational learning versus traditional text-based learning in teaching medical students about flap transplantation,through a randomized controlled trial.Methods In October-November 2025,the Department of Plastic Surgery at Peking University Third Hospital recruited 20 medical students who had not systematically studied flap transplantation knowledge and randomly assigned them equally into either a large language model(LLM)learning group(experimental group,n=10)or a text-based learning group(control group,n=10).All participants studied the same core text on flap transplantation knowledge for 35 minutes.The experimental group learned through conversations with a large language model,while the control group only learned through the text.Knowledge assessments using tests of similar difficulty were conducted before and after the learning ses-sion.Results There were no significant differences in age or pre-test scores between the two groups,indicating comparable baseline levels.Post-test results showed that the LLM group's scores(72.2±13.0)were significantly higher than those of the text-based group(59.8±12.9)(P=0.045).In terms of knowledge improvement,the LLM group(62.0±25.3)％also significantly outperformed the text-based group(21.8±35.6)％(P=0.010).Within-group comparisons revealed that the LLM group showed a highly significant improvement in post-test scores com-pared to pre-test scores(P＜0.0001),whereas the improvement in the text-based group did not reach statistical significance(P=0.117).Con-clusion Within the same learning duration,conversational learning using large language models was significantly more effective than tradi-tional text-based learning in enhancing medical students' short-term mastery of flap transplantation knowledge.This confirms the potential of LLMs as efficient auxiliary tools in specialized medical education.Future research should focus on the development of domain-specific models and evaluate their long-term teaching effectiveness and feasibility in clinical applications.
Keywords：Large language modelsConversational learningPlastic SurgeryFlap transplantationText-based learning
Publication Date：2025-12-15
Online Publishing Date：2026-01-19(First online date of this platform, not the publication date of the document)
Pages：4( 755-758 )