Statistical analysis
Results are presented as percentages (%) and means (±SD) or medians and interquartile range (IQR). Normality was assessed using the Kolmogorov-Smirnov test. Chi-square and Fisher exact tests were used for categorical variables while Student t-test or Mann-Whitney U test were used for quantitative variables as appropriate. Relationships between continuous variables were assessed using Spearman rank correlation coefficient analysis and the correlation coefficient (ρ) was calculated.
Univariate analysis was conducted to assess the factors affecting CRP values in the overall population. Subgroup analysis was performed separately for patients who presented infectious complications. The factors analyzed were VO (vs. Non-VO), ERAS (vs. standard care), male gender (vs.female), CCI >4 (vs. < 4), cancer (vs. benign disease), extent of resection (right hemicolectomyvs. left hemicolectomy vs. rectal resection), need for conversion (vs. no conversion), and surgical time > median (vs. < median). A screened p-value of <0.10 at univariate analysis was then considered for entering the multivariate model after validating the absence of multicollinearity. In multivariate logistic regression, a p value <0.05 was considered statistically significant.
Since POD3 CRP value is used in our clinical practice as one of the discharge criteria, we tested it as predictor of infectious complications by receiver operating characteristic (ROC) curves analysis. The area under the curve (AUC) is a direct measure of diagnostic accuracy of a test. A test with an AUC between 0.7-0.8 is considered as having a good diagnostic accuracy. Cut-off values with the highest sensitivity and specificity were determined by using the Youden’s index. To compare accuracy between ROC curves in subgroup analysis, the DeLong method was applied 14.
Results were considered statistically significant when p value was found to be <0.05. SPSS software (version 23, SPSS, Inc.) was used for statistical analysis. Graphics were outlined using GraphPad Prism version 9.0.