ABSTRACT
Objective:To develop an online interactive cytopathological training program, and to evaluate it for improving the cytopathological diagnostic ability of endoscopists in endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) of pancreas.Methods:A total of 5 500 cytopathological images were collected from 194 patients with pancreatic solid mass who underwent EUS-FNA in Nanjing Drum Tower Hospital from August 2018 to August 2019. The cell type in each cytopathological picture was labeled by senior cellular pathologists, which was used to build a learning and testing platform for online interactive cytopathological training. Five endoscopists without cytopathological background were invited to participate in this training. Sensitivity, specificity, positive predictive value and negative predictive value of endoscopists in differential diagnosis of cancer and non-cancer before and after training were compared to evaluate the effect of the online interactive cytopathological training program on improving the ability of endoscopists in diagnosis of cytopathology.Results:A cytopathological training platform for endoscopists to learn and take online test was successfully built. Before training, sensitivity, specificity, positive predictive value, negative predictive value and accuracy of diagnosis of cancer and non-cancer for endoscopists were 0.55 (95% CI: 0.53-0.58), 0.32 (95% CI: 0.30-0.35), 0.43 (95% CI: 0.41-0.45), 0.44 (95% CI: 0.41-0.47) and 0.43 (95% CI: 0.42-0.45), respectively. After training, the above indicators were 0.96 (95% CI: 0.95-0.97), 0.70 (95% CI: 0.68-0.73), 0.74 (95% CI: 0.72-0.76), 0.95 (95% CI: 0.94-0.96) and 0.81 (95% CI: 0.80-0.83), respectively, which were significantly improved compared with those before ( P<0.001). Conclusion:The online interactive cytopathological training program can improve the understanding and diagnostic ability of endoscopists in pancreatic cytopathology, help to implement rapid on-site evaluation in the process of EUS-FNA, and improve the diagnostic efficiency of EUS-FNA.