Comparation of different malnutrition screening tools according to GLIM criteria in cancer outpatients

Background Many studies have assessed different malnutrition screening tools in oncologic patients. However, very few have been carried out using the new GLIM criteria for malnutrition. The objective of our study is to compare the most recommended screening tools with respect to the new GLIM criteria for malnutrition in cancer patients. Methods Observational, cross-sectional, and single-center study carried out at the Medical Oncology Department at the Lozano Blesa Hospital in Zaragoza. We recruited 165 patients with tumors of the upper-gastrointestinal-tract, colorectal, and head-and-neck region undergoing outpatient treatment. All of them received MST, MUST, Nutriscore, MNA and CONUT screening tools, as well as the GLIM diagnostic criteria, which was used as the gold standard. Results MNA-SF showed the best sensitivity (0.99) and lowest specificity while CONUT had the best specificity (0.89) and lowest sensitivity to detect cancer-related malnutrition. We observed high variability in the diagnostic capabilities of Nutriscore when tumor location was considered, reducing sensitivity in patients with colorectal cancer compared to those with tumors of the upper-gastrointestinal-tract or head-and-neck location (0.25, 0.83, and 0.91 respectively). The highest index of agreement between the screening tools was found between MST, MUST and Nutriscore tests. Regarding the GLIM criteria, the highest agreement index was presented by MUST tool (0.66), while CONUT presented the lowest (0.12). Conclusions Selecting the screening tool according to the type of cancer and its location may allow us to optimize its use and increase its performance, exploiting the advantages of each of them in the different populations.