Comparison of various scoring methods for the diagnosis of protein–energy wasting in hemodialysis patients

Abstract

The present study was designed to determine the cutoff points for the diagnosis of mild-to-moderate and severe protein-energy wasting (PEW) based on dialysis malnutrition score (DMS) and malnutrition inflammation score (MIS), and the sensitivity, specificity, accuracy, area under receiver operating characteristic (ROC) curve, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR(+)) and negative likelihood ratio (LR(-)) of DMS and MIS in comparison with subjective global assessment (SGA) in hemodialysis (HD) patients. In this study, 291 HD patients were randomly selected from among 2,302 adult HD patients in Tehran hemodialysis centers. The PEW in these patients was determined by SGA, DMS and MIS. According to the cutoff points derived from the area under ROC curves, scores of 7-13 for DMS represented normal status or without PEW; 14-23, mild-to-moderate PEW; and 24-35, severe PEW. For MIS, scores of 0-7 represented normal status or without PEW; 8-18, mild-to-moderate PEW; and 19-30, severe PEW. In comparison with SGA, the sensitivity, specificity, accuracy, area under ROC curve, PPV, NPV, LR(+) and LR(-) of DMS were 94 %, 88 %, 92 %, 97 %, 93 %, 92 %, 7.8 and 0.07, respectively. Those of MIS were 87 %, 96 %, 91 %, 97 %, 97 %, 83 %, 22.0 and 0.13 in comparison with SGA, respectively. The results of the present study indicate that the DMS and MIS are almost similar to SGA for identifying PEW in HD patients, but it seems that the DMS is a more appropriate alternative tool for SGA in hospital routine assessments.