Integrating the Colon Leakage Score (CLS) and Serum Albumin to Predict Anastomotic Leakage in Colorectal Cancer Surgery: A Diagnostic Accuracy Study in an Indonesian Cohort

Muhammad Rizqi Firyal (Specialist Doctor Study Program, Faculty of Medicine, Universitas Sriwijaya, Palembang, Indonesia)
Muhammad Hafidh Komar (Department of Digestive Surgery, Faculty of Medicine, Universitas Sriwijaya, Palembang, Indonesia)
Theodorus (Department of Pharmacology, Faculty of Medicine, Universitas Sriwijaya, Palembang, Indonesia)

Sriwijaya Journal of Surgery
Vol. 8 No. 2 (2025): Sriwijaya Journal of Surgery

Introduction: Anastomotic leakage (AL) is a devastating complication in colorectal surgery, associated with high rates of morbidity and mortality. Accurate preoperative risk stratification is essential for guiding clinical decision-making. This study aimed to evaluate the diagnostic accuracy of a synergistic model combining the clinical Colon Leakage Score (CLS) with the biochemical marker of preoperative serum albumin for predicting AL in an Indonesian patient cohort. Methods: A retrospective diagnostic accuracy study was conducted at a single tertiary care center. The study included 60 patients who underwent resection and primary anastomosis for colorectal cancer between January 2022 and June 2024. Patients who received a diverting stoma were excluded. A "high-risk" status was defined by a composite criterion: a CLS > 11 and a preoperative serum albumin level < 3.5 g/dL. The primary outcome was clinically significant AL. Following the identification of inconsistencies in the initial analysis, a complete data re-analysis was performed. Sensitivity, specificity, positive (PPV) and negative (NPV) predictive values, and overall accuracy, with 95% confidence intervals (CI), were calculated. Results: The incidence of AL was 21.7% (13/60 patients). A striking 71.7% of the cohort presented with preoperative hypoalbuminemia. The analysis revealed that the combined model demonstrated poor sensitivity of 46.2% (95% CI: 19.2% - 74.9%) but excellent specificity of 97.9% (95% CI: 88.7% - 99.9%). The model yielded a high PPV of 85.7% (95% CI: 42.1% - 99.6%) and a robust NPV of 86.8% (95% CI: 75.0% - 94.6%). The overall accuracy was 86.7%. Conclusion: The combined CLS-albumin model functions as a highly specific "rule-in" test, not a general screening tool. While it fails to identify more than half of the patients who will leak, a positive result correctly identifies a small subset of patients at extremely high risk for anastomotic leakage. The findings also highlight a profound baseline burden of malnutrition in this population, which warrants further investigation and clinical attention.