External validation of anastomotic leakage risk analysis system in patients who underwent colorectal resection

ABSTRACT

Background/aim: One of the most feared complications after colon resection for carcinoma is anastomotic leakage. Prediction of anastomotic leakage can alter pre- and perioperative management of patients. This study validates an anastomotic leakage prediction system. Materials and methods: Ninety-five patients who underwent colonic resection between 1 January 2016 and 30 January 2017 were included in the study. Patient records and electronic charting system data were used to calculate anastomotic leakage risk on the http//www.anastomoticleak.com/ website. Results: Fifty-six (58.9%) patients were male and thirty-nine (41.1%) were female. The mean age was 61.7 (min 33, max 90). Six (6.3%) patients had anastomotic leakage. According to the ROC analysis, the area under curve for the prediction system was 0.767. Conclusion: The prediction system for anastomotic leakage produced significant results for our patient population. It can be effectively utilized in preoperative and perioperative measures to prevent anastomotic leakage.