Analysis Of Risk Factors For Pneumonia After Esophagectomy And Establishment Of A Risk Predictive Model

Background and ObjectiveEsophageal cancer is one of the most common malignant types in the world.China is the country with the highest incidence and mortality of esophageal cancer in the world.Currently,surgical resection is the main method for the treatment of esophageal cancer.Although the continuous improvement of surgical and postoperative management techniques improved the outcome of esophagectomy,postoperative respiratory complications are still the most common complication after esophagectomy,especially postoperative pneumonia.Postoperative pneumonia is the most common adverse event after esophagectomy.It will prolong hospital stay,increase operative mortality,increase hospitalization costs,and even affect long-term survival and health-related quality of life.With previous reports,the incidence of pneumonia after esophageal cancer was 10%-28.9%.Postoperative pneumonia in esophageal cancer has gradually attracted more attention,and relevant reports have gradually increased in recent years.The risk factors for postoperative pneumonia in esophageal cancer have been reported.However,in the past 20 years,simple intraoperative or postoperative management has not been able to reduce the incidence of postoperative pulmonary complications of esophageal cancer to an ideal level.How to predict the risk of postoperative pneumonia according to the preoperative,intraoperative and postoperative conditions,and take appropriate countermeasures to reduce the incidence of postoperative pneumonia is a hot topic in recent years.Some foreign scholars have developed risk scores for predicting postoperative pulmonary complications based on age,PS scores,FEV1%,and DLCO%.However,the scoring system was developed and validated by patients with similar characteristics and has not been validated outside the United States.In addition,the use of a risk score system for predicting postoperative complications of esophageal cancer to predict the risk of postoperative pneumonia is less accurate.And the risk scoring system for predicting postoperative pneumonia in esophageal cancer is rarely reported.Therefore,this study aims to analyze the occurrence of postoperative pneumonia in patients with esophageal cancer and related risk factors,and use independent risk factors to establish a new prediction model for postoperative pneumonia prediction,which is able to more accurately predict the occurrence of postoperative pneumonia and to provide adequate clinical basis for the patient’s individualized treatment.MethodFrom January 2015 to March 2016,clinical data of 806 patients with primary esophageal cancer diagnosed at the affiliated cancer hospital of Zhengzhou University and undergoing radical surgery for esophageal cancer at thoracic surgery were analyzed retrospectively.Among them,646 patients from January 2015 to December 2015 were used as a model group to use their data to establish a risk prediction score of postoperative pneumonia and to test the risk prediction scores,according to the subject’s operating characteristics(ROC)area under the curve(AUC)judges the effectiveness of the risk prediction score table.From January 2016 to March 2016,160 patients served as verification teams to verify the risk score table and further judge the predictive ability of the risk score table.ResultsUnivariate analysis and multivariate analysis showed that smoking index,surgical methods(thoracic laparoscopy or open),POD1 neutrophil percentage(%),postoperative fasting blood glucose,anastomotic leakage and recurrent laryngeal nerve paralysis were independent risk factors of postoperative pneumonia.According to the results of multivariate logistic regression analysis,a scoring scale for predicting postoperative pneumonia of esophageal cancer was established.This predictive score sheet contains 6 scoring items,with scores ranging from 0 to 10,divided into four risk categories: Grade I(0-2 points),grade II(3-4 points),grade III(5-6 points),and grade IV(7-10 points).The ROC curve plotted had an AUC of 0.721,which was statistically significant compared with 0.5(P<0.001),indicating that this risk score table has good predictive value.The clinical data of patients in the validation group were substituted into the predictive score table,and the ROC curve was plotted.The AUC was 0.736,which was statistically significant compared with 0.5(P<0.001),indicating that this risk score table has good predictive value.Grades I,II,III,and IV were 58.1%,31.9%,6.9%,and 3.1%,respectively,and postoperative pneumonia rates were 6.5%,17.6%,36.4%,and 80.0%,respectively.The difference was statistically significant(P<0.001),indicating that the scale has a higher predictive power for postoperative pneumonia.ConclusionsSmoking index,surgical methods(thoracic laparoscopic or open),POD1 neutrophil percentage(%),postoperative fasting blood glucose,anastomotic leakage,and recurrent laryngeal nerve paralysis are independent risk factors for postoperative pneumonia.The predictive model constructed based on the independent risk factors for postoperative pneumonia can predict the risk of postoperative pneumonia simply and effectively.However,a more reliable prediction score needs the verification of a multicenter,large sample of prospective trials.