Comparison of nutritional diagnosis methods and prediction of clinical outcomes in patients with neoplasms and digestive tract diseases

Abstract

Nowadays studies are advised to compare nutritional risk assessed by different instruments with clinical outcomes. This study compared nutritional diagnosis methods and identified the best predictor of clinical outcomes. This cross-sectional study included 500 hospitalized patients with neoplasms and digestive tract diseases (DTD). Their nutritional status was determined by nutritional risk screening (NRS), subjective global assessment (SGA), and anthropometry, and compared with the clinical outcomes. The Kappa coefficient measured the agreement between the methods. Associations between risk factors and clinical outcomes were investigated by Cox, univariate logistic, and multiple logistic regression analyses at a significance level of 5%. In DTD and cancer patients, SGA and NRS presented good agreement, but agreement of either with anthropometry was poor. According to Cox regression, both SGA and NRS predicted complications in DTD patients. However, none of the instruments was capable of predicting complications in cancer patients or death in DTD patients. In cancer patients, SGA and age were considered risk factors for death. In DTD patients, age, SGA, and NRS predicted a long hospital stay. In cancer patients, long stays were associated with age and SGA. SGA and NRS are highly sensitive for predicting complications in DTD patients. Old age and SGA predicted death in cancer patients. Advanced age and SGA also predicted long hospital stays for DTD and cancer patients, but NRS predicted long hospital stays only for DTD patients.