To develop an aggregate score system to stratify the risk of anastomotic leakage after esophageal carcinoma surgery so as toscreen the patients with high risk for post-operative anastomotic leak.

The clinical data of patients undergoing esophageal carcinoma surgery in Department of Thoracic Surgery, Beijing Chaoyang Hospital between January 2010 and January 2015 were retrospectively collected, and a total of 285 patients were included for analysis. The independent risk factors for anastomotic leak after surgery were determined by univariate analysis and multivariate Logistic regression analysis. The scoring system was developed by proportional weighing of the independent risk factors, and patients were grouped into different classes according to their total score.

The incidence of anastomotic leak after operation was 9.5%(27/285), and the in-hospital mortality was 2.8%(8/285). Multivariate Logistic regression analysis revealed that the site of anastomosis(neck), age (>69 years) and hypoalbuminemia (<25 g/L) were independently correlated with anastomotic leak after esophageal carcinoma surgery. With each risk factor receiving weighted score of 1, the patients were divided into 4 groups with an increased incidence of anastomotic leak: score 0 group, 2.2%; score 1 group, 7.4%; score 2 group, 21.0%; and score 3 group, 37.5%.

The aggregate score system is a reliable and practical tool for identifying patients with high risk of developing anastomotic leak after esophageal carcinoma surgery.