Low sensitivity of commonly used malnutrition screening tools can be explained by the limited detection of isolated...

Growing evidence suggests that malnutrition screening tools show limited agreement with the Global Leadership Initiative on Malnutrition (GLIM) diagnostic reference standard. We aimed to evaluate which GLIM phenotypic criteria are detected by commonly used screening tools in hospitalized patients and identify mismatches.

Secondary analysis of 207 hospitalized patients with GLIM-defined malnutrition. Five screening tools-the Short Nutritional Assessment Questionnaire (SNAQ), Malnutrition Universal Screening Tool (MUST), Malnutrition Screening Tool (MST), Mini Nutritional Assessment-Short Form (MNA-SF), and Patient-Generated Subjective Global Assessment-Short Form (PG-SGA SF) - were assessed for their ability to detect low BMI, reduced muscle mass, and involuntary weight loss. Post hoc analyses (excluding PG-SGA SF) focused on patients with isolated low muscle mass. As nearly all patients fulfilled one or both etiological criteria, the present analysis is primarily directed at the phenotypic criteria.

Sensitivity for GLIM-defined malnutrition was low: MUST 25.1%, MNA-SF 27.1%, SNAQ 46.9%, MST 52.1%, and PG-SGA SF 68.6%. Tools most often identified involuntary weight loss (33.9-76.8%), followed by low BMI (46.5-70.4%), and were least effective for reduced muscle mass (24.8-68.8%). Among malnourished patients with only low muscle mass as phenotypic criterion (n = 61), detection was minimal: SNAQ 3.3%, MUST 0%, MST 8.2%, and MNA-SF 1.6%.

Common screening tools perform poorly in identifying GLIM-defined malnutrition. They mainly detect weight loss, moderately capture low BMI, and mostly miss reduced muscle mass, leaving patients with isolated low muscle mass largely unrecognized.