Objectives: Since the launch of Global Leadership Initiative on Malnutrition (GLIM), there has been an urgent need to validate the new criteria, especially in patients with cancer. The aim of this study was to evaluate and validate the use of the GLIM criteria in patients with cancer.
Method: This multicenter cohort study compared the GLIM with the scored Patient-Generated Subjective Global Assessment (sPG-SGA). The 1-y survival rate, multivariate Cox regression analysis, κ-value, sensitivity, specificity, receiver operating characteristic (ROC) curve, and time-dependent ROC analysis were applied to identify the performance of the GLIM.
Results: Among the 3777 patients in the study, 50.9% versus 49.1% or 36.3% versus 63.7% of the patients were defined as well-nourished and malnourished by GLIM or sPG-SGA, respectively. GLIM presented moderate consistency (κ = 0.54, P < 0.001), fair sensitivity and specificity (70.5 and 88.3%) compared with sPG-SGA. There was no difference in the 1-y survival rate in malnourished patients (76.9 versus 76.4%, P = 0.711), but it was significantly different in well-nourished patients (85.8 versus 90.3%, P < 0.001) between GLIM and sPG-SGA. The above difference was eliminated after omitted nutritional risk screening (NRS)-2002 screening before GLIM (88.1 versus 90.3%, P = 0.078). Omitting NRS-2002 screening before GLIM did not change the 1-y survival rate in well-nourished or malnourished patients by GLIM with NRS-2002 screening (76.9 versus 78.9%, P = 0.099; 85.8% versus 88.1%, P = 0.092) although it significantly raised the rate of malnutrition to 72.5%. The combination of "weight loss and cancer" showed better performance than other combinations.
Conclusions: GLIM could be a convenient alternative to sPG-SGA in nutrition assessment for patients with cancer. The combination of "weight loss and cancer" was better than other combinations. Considering the higher risk for malnutrition in patients with cancer, NRS-2002 screening may not be needed before GLIM.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1016/j.nut.2021.111379 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!