ABSTRACT
INTRODUCTION: The postoperative decease is the most severe complication in surgery. PATIENTS AND METHOD: Within the framework of a multicentre study between January 1 (st) and December 31, 1999, 3,756 patients, 1,463 of them suffering from rectal cancer and 2 293 from colon carcinoma, from 75 clinics were documented with the help of a standardized questionnaire. We compared data of 211 patients who died postoperatively with data of 3,484 patients who survived after surgical treatment of colorectal cancer. Logistical regressions, under inclusion and exclusion of intra- and postoperative complications, show independent influence factors on the postoperative decease and provide models for the prediction of the postoperative death. RESULTS: Compared to the patients who survived, the postoperative deceased patients were significantly older. They had a poorer risk profile and therefore a higher ASA-score (p < 0.001). 20.4 % of the patients underwent an emergency operation. General and specific postoperative complications occurred significantly more frequently. The model of a logistical regression allowed the prediction of postoperative decease with a sensitivity and specificity of 91 %. General postoperative complications such as pulmonary embolism (relative risk: 30.3), cardiac (relative risk: 24.1), renal (relative risk: 22.1), and pulmonary complications (relative risk: 12.0) are crucial for lethality. DISCUSSION: The postoperative decease is influenced by several factors. It is impossible to reduce the number of influence factors for the prediction of outcome. The general postoperative complications, however, represent a crucial problem. It is important to avoid these problems in order to reduce postoperative lethality.
