Factors Affecting Hospital Mortality in Patients with Esophagogastric Anastomotic Leak: A Retrospective Study

We aimed to clarify the association between anastomotic leak and leak-associated mortality to assist decision-making and reduce hospital mortality. Anastomotic leak is a common complication after esophagectomy, but the nature of its relationship to leak-associated mortality has not been established. A retrospective review of all esophagogastric anastomotic leaks that had occurred between 2008 and 2012 at our institution (n = 246) was performed. Risk factors for leak-associated mortality were determined using a multivariate logistic regression analysis. Of the 246 patients with anastomotic leaks, 14 (5.7 %) died. Leak-associated mortality rates were similar regardless of anastomosis location (cervical vs. thoracic anastomosis), surgical approaches (retrosternal vs. prevertebral reconstruction route) and anastomotic techniques (hand-sewn vs. mechanical anastomosis). When a leak occurred, risk factors for leak-associated mortality as determined by multivariate logistic analysis included patient age > 60 years (P = 0.029) and the occurrence of the leak within 1 week of surgery (P = 0.039). When disease worsened after treatment, leak-associated mortality was more frequent in patients requiring reintubation (25.6 vs. 1.4 %, P < 0.001). Fatal bleeding and sepsis were the most common causes of leak-associated mortality. In patients with anastomotic leaks, patient age > 60 years and the occurrence of the leak within 1 week of surgery were risk factors for leak-associated mortality. Increased efforts to reduce the incidence of early anastomotic leaks within 1 week after surgery and prevent the need for reintubation are important for improving patient prognosis.