• all cause anastomotic leakage [ Time Frame: two years ] [ Designated as safety issue: No ]
  

• Time variation of starting oral feeding and hospital stay [ Time Frame: two years ] [ Designated as safety issue: No ]
  

Goal:

The effect of oral hygiene on the occurrence of esophagogastric anastomotic leakage has not yet been studied for along time. We will use a random cohort study model and investigate the effect of perioperative oral anti-infective gargle agent on the esophagogastric anastomotic wound healing.

Method:

1. Design:

   One hundred and twenty patients are divided into 3 groups and each group has 40 esophageal patients. Minimization stratified randomization will be applied. Oral anti-infective gargling agent, Mycostatin oral suspension, will be employed for one week before operation. The first groups will be treated without oral gargle agent; the second groups will be treated with gargling water; and the third groups will be treated with anti-infective gargling agent.

   An end-to-side two-layer esophagogastric anastomosis will be constructed using interrupted sutures with metallic staple through cervical wound. On the other hand, the anastomotic leakage rates in different groups will be investigated.

2. Data Collection and Statistic Analysis:

The records of all patients, various biologic parameters, and the management of leakage are analyzed. Thirty-day morbidity and mortality are determined, and stepwise multivariable logistic regression analysis assesses the effect of preoperative and postoperative variables on anastomotic leakage. Time variation of starting oral feeding and hospital stay are compared using the Kaplan-Meier method.