Abstract: Objective To explore the effect of short-term semi-immersion teaching and training based on the learning pyramid model to improve the capacity building of to find early gastric cancer by digestive endoscope physicians. Methods Thirty digestive refresher doctors were randomly selected from our center for 4-week semi-immersion training. The capacity of the trainees to recognize early gastric cancer was tested before and after 4 weeks of training, and the accuracy of the test was compared. Results The accuracy of the test before training, after theoretical lesson and after 4 weeks of training were 44.58%±16.39%, 48.54%±16.95% and 68.75%±12.06% respectively. The test accuracy after theory lesson plus 4 weeks of training was higher than that before training and after only theoretical lesson (P＜0.001).The test accuracy after theoretical lesson was higher than that before the training(P＜0.05). When grouping by the number of cases in previous gastroscopy, no matter before training or after theoretical lesson, the test accuracy with experience in testing was higher than group with less experience. However, the test accuracy of the two groups almost reached a same level after 4 weeks of training(P＜0.01). With the extension of training time, the test accuracy gradually increased (P＜0.001). There was an interaction between grouping and training time (P＜0.001).In the correlation analysis, the number of cases of the previous gastroscopy performance was positively correlated with the test accuracy before training and after theoretical lesson (r=0.668,＜0.001,r=0.581, P＜0.01). Conclusions Semi-immersion training guided by the learning pyramid model can effectively improve the capacity of digestive endoscope physicians to recognize early gastric cancer.

Key words: diagnosis capacity of early gastric cancer, learning pyramid model theory, semi-immersion teaching training