NUTRISCORE Compared to NRS-2002 for Detection of Malnutrition in Hospitalized Cancer Patients; A Cross Sectional Study

Abstract

The most commonly recommended screening test for determining the nutritional status of hospitalized cancer patients is Nutrition Risk Screening-2002 (NRS-2002). NUTRISCORE, on the other hand, is an outpatient cancer patient-specific nutritional screening test that is easier to administer than NRS-2002 and queries tumor location and treatment information received from the patient. We aimed to investigate the validity of NUTRISCORE, in hospitalized cancer patients. In total, 112 patients were enrolled in this study. The NRS 2002 and NUTRISCORE screening tests were performed. The data obtained with NUTRISCORE were compared to the reference test NRS-2002 using the κ test and ROC curve analysis. The NRS-2002 identified 45.5% of patients as being at risk of malnutrition, while the NUTRISCORE test identified 48.2% (k = 0.516, p < 0.005). The AUC value was 0.759 (95% CI:0.67-0.85) as shown in the ROC analysis. Using the NRS-2002 as a reference test, the sensitivity (S), specificity (SP), Positive Predictive (PPD), and Negative Predictive (NPD) values for the NUTRISCORE test were 76.5% (95% CI:63.7-86.6), 75.4% (95.CI:63.7-85), 72.2% (95% CI:59.4-83), 79%, (95% CI:67.7-88.3) respectively. NUTRISCORE can be used for screening of malnutrition in hospitalized cancer patients.