Complementaridade de diferentes ferramentas de triagem de risco nutricional aos critérios GLIM para diagnóstico de desnutrição em pacientes com câncer

Abstract

Introduction: Cancer is the second leading cause of death worldwide and the course of the disease increases the risk of malnutrition. Approximately 20% of patients die from malnutrition and associated complications, not from the disease itself. Therefore, the correct identification of malnutrition favors early intervention in those who need it. Nutritional risk screening (NRS) and nutritional diagnosis tools become essential in this process. Objectives: To evaluate the complementarity of different NRT tools to the GLIM criteria for diagnosing malnutrition in cancer patients. Methods: The study represents a secondary analysis of a prospective cohort, which evaluated cancer patients in a hospital setting. The NRS was made using the tools Protocol for Nutritional Risk in Oncology (PRONTO), Patient-Generated Subjective Global Assessment (PG-SGA), Nutritional Risk Screening 2002 (NRS-2002), Malnutrition Screening Tool (MST), Malnutrition Universal Screening Tool (MUST) and Nutriscore. The assessment of nutritional status was carried out using the criteria proposed by GLIM, considering phenotypic criteria of weight loss, reduced Body Mass Index and low muscle mass (via calf circumference and fat-free mass index assessed by bioelectrical impedance), combined with etiology criteria of reduced food intake and inflammation, considered present in all patients due to cancer diagnosis. Patients were evaluated up to 12 months after hospital discharge to assess the outcome of interest (death). To verify the complementarity of NRS tools to the GLIM criteria for diagnosing malnutrition, accuracy metrics were calculated and Cox regression analysis was conducted to verify the association between mortality and the presence of nutritional risk combined with a diagnosis of malnutrition, considering the different tools of NRS. Results: 290 patients were evaluated, the majority of whom were elderly (60.3%), female (53.1%) and with solid tumors (89.7%). Of the total, 118 (40.7%) individuals died. The PRONTO tool identified more individuals (n = 240, 82.8%) at nutritional risk. However, like the others, it presented reduced complementarity to the GLIM criteria, taking into account the specificity (68.6%) and negative predictive value (48.0%) obtained. Regarding mortality risk, after adjustment for confounding variables, the presence of malnutrition according to GLIM, without prior screening, was associated with mortality (HR: 5.15; 95%CI 1.62 – 16.36; p < 0.001). Nutritional risk was also a predictor of death, regardless of the NRS tool used (HR between 1.70 and 10.07). Furthermore, the presence of nutritional risk and malnutrition, regardless of the NRS tool used, increased the risk of death between 1.90 and 7.70 times. Conclusion: The prevalence of malnutrition according to the GLIM criteria is dependent on the tool applied for NRT prior to assessing the presence of etiological and phenotypic criteria, and was higher when applied to ASG-PPP. No NRS tool presented satisfactory specificity and negative predictive value values. Nutritional risk and malnutrition, isolated or combined, regardless of the tool applied, were predictors of death in cancer patients. These results suggest that, regardless of the NRS tool used and the result, the diagnosis of malnutrition using the GLIM criteria should be performed in all cancer patients.