Establishment And Validation Of Prediction Model For Pulmonary Complications After Total Gastrectomy For Upper Gastric Cancer

Object:Gastric cancer is one of the diseases with the highest morbidity and mortality in the world,among which the incidence of upper gastric cancer is increasing year by year.The upper gastric cancer has been proved to have unique course and prognosis.Surgery is still the main treatment,with total gastrectomy being the main operation.The occurrence of postoperative pulmonary complications has an important impact on the postoperative recovery and disease diagnosis and treatment of patients with upper gastric cancer.The risk factors related to pulmonary complications after total gastrectomy for upper gastric cancer were clarified,and the risk prediction model of the nomogram was established.It provides some basis for the effective prevention and reduction of pulmonary complications in clinical work.Methods:Patients who underwent total gastrectomy for upper gastric cancer in Affiliated Hospital of Qingdao University from October 2019 to September 2021 were collected and screened for inclusion in the study in strict accordance of inclusion and exclusion criteria,and the postoperative pulmonary complications were statistically analyzed.All patients were randomly divided into modeling group and validation group in a ratio of 7:3 using random number table method.Univariate and further multivariate logistic analysis were carried out in the modeling group,and finally the nomogram was constructed.Validation group was used to verify the accuracy and practicability of the model.The receiver operating characteristic curve and area under the curve,calibration curve,C-index and decision curve analysis were used to evaluate the model.Results:The 280 patients included were divided into 7:3 groups,including 196 in the modeling group and 84 in the verification group.Among the overall data,118 patients developed pulmonary complications,with a rate of 42.1%.In the modeling group,77 cases(39.2%)developed pulmonary complications.Verification group 41 cases,the incidence of48.8%.Univariate analysis based on the data of the modeling group showed that smoking duration,age,preoperative hypertension,diabetes,preoperative pulmonary complications,smoking history,intraoperative blood transfusion,BMI>26kg/m~2,preoperative albumin<35g/L were correlated with postoperative pulmonary complications.Multiple logistic regression analysis showed that age,smoking history,preoperative diabetes mellitus,preoperative albumin<35g/L,BMI>26kg/m~2 were independent risk factors for pulmonary complications after total gastrectomy for upper gastric cancer.On this basis,a nomogram prediction model was constructed.The area under ROC curve of the modeling group and the verification group were 0.809 and 0.767,respectively.Together with C-index,this prediction model is proved to be effective.The two groups of calibration curves showed that the predicted probability of pulmonary complications was consistent with the actual probability.The analysis of decision curve shows that the constructed model has clinical practicability.Conclusion:Age,smoking history,preoperative albumin<35g/L,preoperative diabetes mellitus,BMI>26kg/m~2 were independent risk factors for pulmonary complications after total gastrectomy for upper gastric cancer.The prediction model constructed by the nomogram could individually predict the probability of postoperative pulmonary complications in specific patients with upper gastric cancer.It can also guide front-line medical staff to pay specific attention to perioperative indicators of patients and take corresponding measures before,during and after surgery,which is conducive to reducing the incidence of postoperative pulmonary complications of upper gastric cancer and improving the quality of diagnosis and treatment.