Nutritional risk screening 2002 (NRS2002) is a commonly used tool for screening the risk of malnutrition in hospitalized patients, while patient-generated subjective global assessment (PG-SGA) is a nutritional assessment tool for malignant tumor patients. However, there are still gaps in the rapid nutritional risk screening methods for cancer patients. We aimed to evaluate the value of abridged scored patient-generated subjective global assessment (abPG-SGA) for nutritional risk screening and prognosis in cancer patients. The NRS 2002 and abPG-SGA scores of 100 malignant tumor patients hospitalized in our department in December 2020 were collected. Take NRS2002 ≥ 3 as the positive standard (risk of malnutrition). Data were analyzed using Kappa test, ROC curves, cut-off values and Kaplan-Meier. In the screening of 100 patients, 25.0% of patients were at risk of malnutrition (NRS2002), abPG-SGA yielded a sensitivity and specificity of 92.0% and 72.0%, respectively (area under curve [AUC] = 0.884, cut-off value ≥ 4.5); In the screening of patients with digestive system malignancies, 22.6% of patients were at risk of malnutrition (NRS2002), and the sensitivity and specificity of abPG-SGA were 91.67% and 87.80%, respectively (AUC = 0.945, cut-off value ≥ 5.5). The results of survival analysis showed that the overall survival (OS) of patients with abPG-SGA ≥ 5 and < 5, NRS2002 ≥ 3 and abPG-SGA < 5, NRS2002 < 3 and abPG-SGA ≥ 5 were significantly different (P < .0001), the OS of patients with NRS2002 ≥ 3 and abPG-SGA ≥ 5, NRS2002 < 3 and abPG-SGA < 5 were not significantly different (P > .05). Like NRS2002, abPG-SGA can also be used for malnutrition screening and prognosis judgment in cancer patients. It can quickly screen out cancer patients who may be at risk of malnutrition and facilitate the development of nutritional assessments.
Read full abstract