Statistical analysis

No power analysis was performed as the incidence of PPLOS in the study population has never been reported in the literature.

Continuous variables are described as mean and standard deviation or as median (25th, 75th percentiles) and were compared using an independent t test or the Mann–Whitney U test, as appropriate. Categorical variables are reported as numbers (proportions) and were compared using the Fisher exact test.

Variables considered clinically relevant or variables with a P-value < 0.1 in univariate analysis were introduced into the multivariate logistic regression model. Numerical or multi-categorical variables were transformed into two- or three-category variables before being entered into the multivariable logistic regression model. For example, we combined similar surgical procedures (e.g., anterior rectal cancer resection [Dixon] and sigmoidectomy) or the procedures with similar trends for PPLOS (e.g., coloanal anastomosis [Parks] and right hemicolectomy) into the same category to form a three-category variable. Variables with missing data were not included in multivariate analysis. Once a final list of variables was constructed, a stepwise logistic regression analysis was performed to examine the relationship between potential influencing factors and PPLOS. The final multivariate model was tested for goodness of fit using the Hosmer–Lemeshow test. The identified influencing factors are presented as odds ratios (ORs) with 95% confidence intervals (CIs).

After identifying the influencing factors for PPLOS, we conducted a further analysis of the association between clinical characteristics and postoperative albumin infusion. Additionally, we conducted a stratified univariate logistic regression analysis (stratification factors: preoperative albumin level and duration of surgery) of the association between postoperative albumin infusion and PPLOS.

Statistical analyses were conducted using SPSS Statistics for Windows, version 17.0 (SPSS Inc., Chicago, IL, USA). All P-values were two-sided, and a P-value < 0.05 was considered statistically significant.