Objective To evaluate the preoperative nutritional status of patients with gastric carcinoma by using the European Nutritional Risk Screening 2002(NRS 2002)and its prediction for postoperative nutrition-related complications.Methods We prospectively evaluated the nutritional risk of 314 gastric cancer patients admitted in one center from 2004 to 2007 with NRS 2002 with China's normal body mass index(BMI),in terms of postoperative complications,mortality and hospital stay.Results NRS 2002 scoring system was applicable in 93.1% cases.Preoperatively 125 patients were of score≥3,accounting for 39.8% of this group.The postoperative complication rate(26.2%)was higher than 13.8% in those with normal preoperative nutritional scores(NRS 2002 score＜3)(P＜0.05);The odds ratio to develop a complication was 0.642 in patients with preoperative nutritional risk score(P＜0.05),and 1.596 in patients with clinicopathological stage of gastric cancer(P＜0.01).The correlation between length of hospital stay and nutritional risk was also assessed by Pearson correlation analysis.The Pearson coefficient was 0.177(P=0.002).Conclusion Preoperative nutrition score(NRS 2002)≥3 predicts higher postoperative complications and longer hospital stay.Preoperative nutritional support is necessary in patients with preoperative nutrition score(NRS 2002)≥3.

Objective To investigate the clinical value of nutritional risk screening 2002(NRS2002)and malnutritional universal screening tools(MUST)in the preoperative nutrition risk evaluation of patients with gastric cancer.Methods The preoperative nutritional risk of 3 14 patients who had been admitted to the Third Affiliated Hospital of Sun Yat-sen University from January 2004 to December 2007 was assessed by subjective global assessment(SGA),NRS 2002 and M UST,and the influence of nutritional risk on the incidence of postoperative complications and hospital stay was investigated.All data were analyzed by Wilcoxon test,Kappa test and Logistics regression analysis.Results Compared with SGA,the sensitivity,specificity,positive predicting value and negative predicting value were 86.7%,74.2%,86.9% and 73.8% for NRS2002,and were 73.1%,70.6%,74.8% and 68.7% for MUST.Compared with MUST,NRS2002 had a higher consistency with SGA(K_(NRS2002)=0.601,K_(MUST)=0.436).Logistic regression analysis revealed that patients with higher MUST or NRS2002 score had higher incidence of postoperative complications and longer hospital stay.In the aspect of hospital stay,the relative risk of MUST was 2.517,which was lower than 3.426 of NRS2002.The relative risk of MUST was 0.529,which was lower than 0.642 of NRS2002 in the aspect of incidence of postoperative complications.Conclusions NRS2002 and MUST are suitable for preoperative nutritional risk screening of patients with gastric cancer,and the score of NRS2002 or MUST is associated with the incidence of postoperative complications and length of hospital stay.However,NRS2002 is more accurate than MUST in the reflection of nutritional risk of patients with gastric cancer.