BACKGROUND: The literature on oral intake after esophagectomy and its influence on anastomotic leakage and complications is sparse.

METHODS: This retrospective study included 359 patients undergoing esophagectomy between January 2011 and August 2015. Three oral intake protocols were evaluated: regimen 1, nil by mouth until postoperative day (POD) 7 followed by a normal diet; regimen 2, oral intake of clear fluids from POD 1 followed by a normal diet; regimen 3, nil by mouth until POD 7 followed by a slow increase to a blended diet. The outcome endpoints were: (1) anastomotic leakage, (2) complications [severity and number described using the Dindo-Clavien Classification and Comprehensive Complication Index (CCI)] and (3) length of stay. A multivariate logistic regression model was obtained for CCI and anastomotic leakage using Wald's stepwise selection.

RESULTS: CCI was significantly lower in regimen 3 (16 vs. 22 and 26 in regimen 1 and 2, p = 0.027). Additionally, significantly fewer patients in regimen 3 suffered from severe complications of Dindo-Clavien grade IIIb-IV (p = 0.025). The incidence of anastomotic leakage reached its lowest in regimen 3, 2%, compared to 7-9%. Multivariate analyses revealed that high American Society of Anesthesiologist score was a predicting factor for both CCI and anastomotic leakage.

CONCLUSION: The study indicates that nil by mouth until postoperative day 7 followed by a slow increase to a blended diet after esophagectomy results in less severe complications and a tendency of fewer anastomotic leakages. Multiple comorbidities proved to be an important predictive factor of the postoperative course.