Maarten van Egmond

26 Chapter 2 of surgical complications. This 5-scale classification reports surgical complications based on the type of therapy required to treat this complication and is an objective, valid, reliable and reproducible classification system. 27,28 We defined POC according to the basic complications list of ECCG, to allow for future comparisons of different surgical approaches and other interventions for esophageal cancer. 26 This complications list distinguishes pulmonary, cardiac, gastrointestinal, urologic, thromboembolic, neurologic/ psychiatric, infection, wound/ diaphragm and other complications. Statistical analysis We summarized patient characteristics with descriptive statistics and expressed discrete variables as counts with percentages. Continuous variables were described as mean and standard deviation and in case of a skewed distribution as median (P50) and interquartile range (P25 to P75). We used univariate logistic regression analysis to assess the association of both conventional risk factors and indicators of functional status with POC. Given the small dataset and to avoid missing true predictors, conventional risk factors and indicators of functional status were entered in a multivariate regression model if P -value ≤ .200. We compared the group of patients suffering from POC with the group of patients without POC on aspects of preoperative functional status by using the independent sample T-test or Mann Whitney U-test. We used backward elimination to create a final multivariate model containing variables with a P -value ≤ .050.