Objectives: The study aimed to evaluate the validity and reliability of the GLIM diagnostic criteria at Hanoi Medical University Hospital.
Methods: A cross-sectional study was conducted on 200 patients aged 18 to under 70 years who were admitted within 48 hours.
Results: Malnourished according to GLIM had significantly longer hospital stays (10.0 ± 8.96 days) and lower clinical indicators, including BMI, subcutaneous fat thickness, body fat percentage, muscle mass, bone mass, visceral fat, and prealbumin levels (p < 0.05). The GLIM criteria demonstrated a sensitivity of 71.9%, a specificity of 88.2%, a positive predictive value of 74.2%, a negative predictive value of 87%, and an area under the receiver operating characteristic curve (AUC) of 0.8 (95%CI = 0.74-0.86) for malnutrition diagnosis. Severe malnutrition according to the GLIM criteria, based on the phenotypic standard at a BMI threshold of < 17.0 kg/m² (for individuals < 70 years old), has a sensitivity of 70%, a specificity of 98.4%, and an AUC of 0.84. The inter-rater reliability between two independent GLIM evaluators was excellent, with a Kappa coefficient of 0.89.
Conclusion: The GLIM criteria demonstrate high validity and reliability in diagnosing malnutrition.