ABSTRACT
OBJECTIVES: The aims of this study were to examine the diagnostic accuracy of the Malnutrition Universal Screening Tool (MUST) and the Patient-Generated Subjective Global Assessment Short Form (PG-SGA-SF) for detecting malnutrition in chronic kidney disease (CKD), study individual contributions of MUST and PG-SGA screening items to the explained variance in nutritional status (NS), and examine whether the PG-SGA-SF score, in combination with one of the items of the clinician's part of the cPG-SGA, can be used as a valid and compact nutrition assessment tool in patients with CKD. METHODS: This was a cross-sectional observational study with 123 patients with CKD who were screened for malnutrition risk by MUST and PG-SGA-SF. NS was determined by complete PG-SGA. Overall accuracy was calculated by the receiver operating curve area under the curve (ROC-AUC). Explained variance of individual screening items was assessed by Nagelkerke's R2, total explained variance was assessed by the increase of R2 after addition of items in manual stepwise forward selection. RESULTS: Of the patients, 44% were malnourished, which was detected by MUST in 24% and by PG-SGA-SF in 78%. Items "body mass index (BMI)" and "no food intake" of the MUST together explained only 3.7% of the variance in NS, whereas the item "nutrition impact symptoms" (NIS) of the PG-SGA-SF explained 57%. Total explained variance in NS by MUST and PG-SGA-SF were 15% and 74%, respectively. The PG-SGA-SF combined with the "metabolic stress" item explained most (87%) and had a sensitivity of 94% to detect malnutrition. CONCLUSIONS: Most malnourished patients with CKD failed to be identified with the MUST, whereas the PG-SGA-SF detected the majority of them with the screening item "NIS" having the highest individual contribution to the explained variance in NS. Combination of PG-SGA-SF with the item "metabolic stress" had the highest overall accuracy to detect malnutrition.
