Predictive Value of the C-Reactive Protein-to-Albumin Ratio for Anastomotic Leakage Following Colorectal Surgery

Background: Anastomotic leakage (AL) is a devastating complication of colorectal surgery, necessitating early and accurate prediction strategies. We investigated the predictive value of the C-reactive protein-to-albumin ratio (CAR) calculated using preoperative albumin levels and a derived bedside leak score for AL prediction. Methods: We conducted a prospective observational study of 72 patients who underwent elective open colorectal surgery with anastomosis between January 2023 and January 2024. We measured the preoperative albumin and C-reactive protein (CRP) levels on postoperative days (POD) 1, 3, and 5. CAR was calculated for POD3 (CAR POD3) and POD5 (CAR POD5) using the preoperative albumin levels. The bedside leakage score was also assessed. The primary outcome was AL within 30 days. We employed Receiver Operating Characteristic (ROC) curve analysis to determine predictive accuracy. Results: AL occurred in 9 of the 72 patients (12.5%). Within this cohort, CAR POD5 demonstrated exceptional discrimination, with an Area Under the Curve (AUC) of 1.000 (95% CI: 1.000-1.000); an optimal cutoff of >46.52 yielded 100% sensitivity and specificity. CAR POD3 also exhibited excellent performance (AUC 0.986, 95% CI: 0.961-1.000; optimal cutoff >15.18, sensitivity 88.9%, specificity 93.7%). In contrast, the Bedside Leak Score performed poorly (AUC 0.447). Lower preoperative albumin levels, reduced body mass index (BMI), and elevated preoperative CRP levels were significantly associated with AL (p =0.001, p =0.002, and p <0.001, respectively). All four (5.6%) postoperative deaths occurred in patients with AL. Conclusions: Postoperative CAR, particularly when assessed on POD5 using preoperative albumin, demonstrated exceptional predictive accuracy for anastomotic leakage after open colorectal surgery within this cohort. CAR on POD3 also provided excellent predictive value. These findings suggest CAR's potential of CAR as a simple and valuable tool for early risk stratification, although its validation in larger multicenter studies is warranted. The Bedside Leak Score did not demonstrate a predictive utility in this study. Clinical trial number: Not applicable.