A dynamic online nomogram based on the Global Leadership Initiative on Malnutrition criteria for predicting 30-day mortality in patients with severe stroke.

Background: The Global Leadership Initiative on Malnutrition (GLIM) has released new criteria for diagnosing and grading malnutrition. However, their validation in patients with severe stroke has not been thoroughly documented. This study aims to investigate the application of GLIM criteria in nutrition assessment and mortality prediction among patients with severe stroke.

Methods: This prospective study included 281 patients with severe stroke from our hospital (July 2021 to July 2023) and a validation set of 121 patients (August 2023 to March 2024). Nutrition status was assessed using Subjective Global Assessment (SGA) and GLIM criteria. The association between nutrition status and 30-day mortality was analyzed using the Kaplan-Meier method and a Cox model. A dynamic nomogram incorporating GLIM and other prognostic variables was developed. The Concordance Index, receiver operating characteristic curve, and calibration curve were used to evaluate predictive accuracy.

Results: GLIM-defined malnutrition was found in 19.93% of the development set and 20.66% of the validation set. Using SGA as a reference tool to evaluate the diagnostic agreement of the GLIM criteria, the GLIM criteria demonstrated good agreement in identifying malnutrition, with a kappa value of 0.669. Patients with GLIM-defined moderate (hazard ratio [HR]: 2.526, P < 0.001) and severe malnutrition (HR: 5.626, P < 0.001) had significantly higher mortality risk than those identified by SGA. The nomogram accurately predicted 30-day mortality.

Conclusions: The GLIM criteria effectively assess malnutrition and predict 30-day mortality in patients with severe stroke. The developed dynamic nomogram provides accurate prognosis prediction.