Statistical Analysis
Continuous data are presented as mean (± standard deviation) for normally distributed variables or median [interquartile range (IQR)] for non-normally distributed variables. Categorical data are displayed as number (percentage). Kaplan Meier analysis was used to evaluate five-year survival and cumulative incidence of all-cause readmission.
Cox proportional hazards modeling was used to model postoperative mortality. In this model, all baseline characteristics and risk factors were assessed in a univariable model. Those with significant associations with mortality (P<0.05) were considered for inclusion in the final multivariable model. Backwards, stepwise elimination was performed to create the final model with covariable inclusion of P<0.2 into the final model. Significant covariable interactions and multicollinearity were investigated.
Competing risk regression was used to model all-cause hospital readmission. In this model, death was the competing event. Baseline characteristics and risk factors were assessed in a univariable model, and those with significant associations (P<0.05) with all-cause readmissions were included in the final multivariable model. Statistical analyses were performed using SAS version 9.2 software (SAS Institute, Cary, NC).