RESUMEN
<p><b>OBJECTIVE</b>To investigate the potential benefits of postoperative nutrition in malnourished patients with gastrointestinal cancer.</p><p><b>METHODS</b>A total of 646 malnourished patients with gastrointestinal cancer defined by the subjective global assessment (SGA) were randomly divided into parenteral nutrition group (n=215), enteral nutrition group (n=215) and conventional group (n=216). Two nutritional regimens were designed to be isocaloric 125.5 kJ(30 kcal).kg(-1).d(-1) and isonitrogenous 0.25 g.kg(-1).d(-1) for 7 postoperative days. Conventional group did not receive artificial nutrition before and after surgery. Postoperative complications, mortality and postoperative length of hospital stay were compared.</p><p><b>RESULTS</b>All baseline and surgical characteristics were comparable among 3 groups. Overall postoperative mortality was 1.5%, and no difference was observed among 3 groups. Postoperative complications occurred in 61(28.4%) patients in enteral nutrition group, 72(33.5%) in parenteral nutrition group, and 97 (44.9%) in conventional group (P=0.000 vs enteral nutrition group; P=0.001 vs parenteral nutrition group). Postoperative length of hospital stay was (9.8+/-3.4) d in enteral nutrition group, (11.2+/-5.0) d in parenteral nutrition group, and (14.5+/-7.1) d in conventional group (P=0.001 vs enteral nutrition group; P=0.003 vs parenteral nutrition group).</p><p><b>CONCLUSIONS</b>Postoperative artificial nutrition support is beneficial to the malnourished patients with gastrointestinal cancer, which improves postoperative outcome. Early enteral nutrition significantly reduces the infectious complication rate and length of postoperative hospital stay as compared with parenteral nutrition.</p>