Ridge regression analysis of pulmonary infection rate after transthoracic esophagectomy for esophageal cancer

Objective To identify the influencing factors of pulmonary infections after transthoracic esophagectomy for esophageal cancer.Methods A retrospective review of 349 patients undergoing transthoracic esophagectomy at our hospital for esophageal cancer was performed between January and December 2009.The postoperative pneumonia rate was examined and 26 perioperative factors possibly affecting the postoperative respiratory complications were collected.Ridge regession modeling was performed to determine if a significant association existed between perioperative factors and postoperative pneumonia.Results The postoperative pneumonia rate of all patients was 27.8％.Eight perioperative factors were found to have affected significantly the postoperative respiratory complications.The influencing factors included (according to contribution):patient control epidural analgesia (PCEA),diabetes,general anesthesia plus epidural analgesia,other postoperative complications,one lung ventilation (OLV),transfusion volume of red blood cells (RBC),body mass index ( BMI ) and age.Conclusion The major influencing factors of pulmonary infection after transthoracic esophagectomy for esophageal cancer are PCEA,diabetes,general anesthesia plus epidural analgesia,other postoperative complications,OLV,transfusion volume of RBC,BMI and age. Key words: Esophageal cancer; Transthoracic esophagectomy; Postoperative pneumonia; Ridge regression