Objective To investigate the related factors and risk model prediction of pneumonia in patients with early gastric cancer after ESD.

Methods We enrolled 171 patients with early gastric cancer who underwent endoscopic submucosal dissection (ESD) from May 2016 to February 2019. Logistic regression analysis was used to screen independent risk factor of pneumonia in patients with early gastric cancer after ESD. Then, based on the selected independent risk factors, we established a nomogram prediction model and verified the accuracy and predictability of the model.

Results Hyperlipidemia, antibacterial usage, ICU treatment or not, swallowing disorder, infiltration depth, oral care and indwelling gastric tube were independent risk factors for pneumonia after ESD. Based on the selected seven independent risk factors, we established a nomogram model to predict the risk of pneumonia after ESD. The validation of the model showed that the predicted values and the observed values were basically consistent, indicating that the nomogram prediction model had a good predictive ability. We used the Bootstrap internal verification method to validate the model. The C-index index was 0.937 (95%CI: 0.929-0.945), indicating that the model had good precision and discrimination.

Conclusion For patients with early gastric cancer, timely consideration of hyperlipidemia, antibiotic usage, ICU treatment or not, dysphagia, depth of infiltration, oral care and indwelling gastric tube could improve the diagnostic efficacy for the incidence of pneumonia after ESD.