Description
Prospective randomised clinical trials of perioperative nutritional support (NS) are reviewed. The following conclusions are drawn. Enteral is preferable to parenteral nutrition, and via nasojejunal tube rather than jejunostomy. Preoperative NS is indicated only in malnourished patients, and the Subjective Global Assessment is an accurate test for malnutrition. Five to ten days of preoperative parenteral NS may decrease postoperative morbidity, while preoperative enteral NS needs further evaluation. Conversely, postoperative parenteral NS may increase postoperative morbidity, while postoperative enteral NS needs further evaluation. Postoperative parenteral NS may however be indicated in those patients who cannot be fed enterally after surgery. Supplementary sip feeds given for one to two weeks following surgery may be beneficial in malnourished patients.

