Aim Our purpose was to investigate the potential role of albumin variation in comparison to C-reactive protein (CRP) variation as a predictive marker for postoperative complications in colorectal surgery.
Methods An prospective cohort study was conducted. Adult patients who underwent elective colorectal surgery between January 2019 and December 2020 were eligible. Serum levels of albumin and CRP were measured preoperatively and on the first 4 postoperative days. Univariate analysis were performed to assess the association of albumin (Alb) and CRP with postoperative complications. Serum albumin variation (ΔAlb) and CRP variation (ΔCRP) were calculated. Receiver operating characteristic curve analysis and the Youden test were used to determine acuity and predictive cut-off values.
Results Ninety-three patients were included. A CRP cut-off of 83.4 mg/dL on postoperative day (POD) 4 was the best predictor of postoperative global complications (p<0.001; AUC 0.83, 70% sensitivity, 91% specificity). Major complications were best correlated with ΔAlb on POD 2, 3 and 4 (p<0.001), with a ΔAlb cut-off of 27.4% on POD 2 showing the strongest association with this outcome (AUC 0.834, 83% sensitivity, 90% specificity). Regarding anastomotic leak, CRP on POD 3 showed better predictive values (p=0.037; AUC 0.792) with a cut-off value of 88.7 mg/dL (100% sensitivity, 52% specificity).
Discussion Herein, the authors demonstrate there is a role for albumin variation, as an earlier and sensitive marker, to predict major postoperative complications in colorectal surgery. This analysis may be further applied to aid in the early identification of significant causes of re-operation and long-term morbimortality.