OBJECTIVEWe retrospectively analyzed the cause and death risk of 114 postoperative respiratory failure patients found in 3519 patients with esophageal cancer and 1495 patients with carcinoma of gastric cardia surgically treated between January 1992 and May 2003.

METHODSTo analyze the reasons causing postoperative respiratory failure in surgically treated esophageal or gastric cardia cancer patients, and the correlation between the death risk of postoperative respiratory failure and preoperative pulmonary function tests, postoperative complications, operation modes, history of preoperative accompanying diseases and so on using Binary Logistic Regression analysis and Chi-square tests (chi(2)) in SSPS statistics software.

RESULTSIn this series, postoperative respiratory failure developed in 97 of 3519 (2.76%) esophageal cancer patients and 17 of 1495 (1.14%) gastric cardia cancer patients, which were mainly caused by severe respiratory tract infection (37.7%, 43/114) and operative complications (35.1%, 40/114) such as: anastomotic leakage or perforation of thoracic stomach, extensive bleeding during operation, chylothorax, etc, totally accounting for 72.8% (83/114). In contrast with lung cancer patients, most of the postoperative respiratory failure (69.3%) occurred in the patients who had perioperative complications but almost always normal preoperative pulmonary function tests. Other reasons to cause postoperative respiratory failure were: extubation in unconscious patients at the end of general anesthesia; over-infusion during operation; pulmonary artery embolism; severe arrhythmia and so on. All patients except 2 were treated in ICU by mechanic ventilation through intubation and/or tracheotomy. Eighty patients (70.2%) were intubated and/or had tracheotomy within 3 days postoperatively. Seventy patients (61.4%) were rescued successfully, whereas 44 cases (38.6%) died of postoperative respiratory failure and/or other postoperative complications. Univariate analysis and multivariate analysis by binary logistic regression indicated that: severe perioperative complications, more postoperative complications, poor preoperative pulmonary function, radical preoperative radiotherapy, intubation and/or tracheotomy after the second postoperative day and long period of mechanic ventilation were the major risk factors leading to death once the postoperative respiratory failure developed. The former 3 factors were independent risk factors leading to death with OR of 2.50, 2.37, 1.68, respectively. Age, sex, operation modes, history of preoperative accompanying disease, prophylactic antibiotics were not demonstrated as statistically significant risk factors correlated with death.

CONCLUSIONSevere perioperative complications and respiratory tract infection are the two major causes of postoperative respiratory failure in patients with cancer of esophagus and gastric cardia. Patients with severe perioperative complications or poor preoperative pulmonary function or association with more than two kinds of postoperative complications have much higher death risk than other patients when they develop postoperative respiratory failure. Careful manipulation during operation and effective perioperative management are the most important measures to avoid postoperative respiratory failure and high mortality.