Performance of the new nutrition evaluation tool for hospitalized pediatric patients with cancer in Brazil (ANPEDCancer).

Background: There is a lack of specific nutrition assessment tools for pediatric patients with cancer. The aim was to evaluate the performance of the ANPEDCancer assessment tool in a pediatric population with cancer, verifying its ability to detect nutrition inadequacy and predict the length of hospital stay (LOS).

Methods: Evaluated 111 pediatric patients hospitalized in the National Cancer Institute (INCA) in 2019 to assess nutrition status. Patients were classified as malnourished and well nourished by the ANPEDCancer. Measures of weight, height, anthropometric indicators, body composition, laboratory parameters, LOS, and death were compared between groups. The ANPEDCancer classification was compared with the complete nutrition assessment, calculating sensitivity, specificity, and predictive values, and with the LOS.

Results: The prevalence of malnutrition was 12.6%, nutrition risk was 48.6%, risk of overweight/obesity was 6.3%, and well-nourished status was 32.4%. According to ANPEDCancer, malnourished patients showed a higher frequency of inadequacy for all anthropometric indices, percentage of weight loss, serum albumin level, C-reactive protein (CRP), and longer LOS when compared with well-nourished patients. There was an association between the tool's diagnosis and measures of body composition, CRP, and LOS. ANPEDCancer validation with the complete nutrition assessment showed a sensitivity of 81.6%, specificity of 55%, positive predictive value of 53.4%, and negative of 82.5%. The LOS was almost twice as long among malnourished patients and was statistically significant (P = 0.002).

Conclusion: ANPEDCancer is a feasible tool to assess nutrition status and identify the presence of nutrition risk, allowing for targeted assistance in hospitalized pediatric patients with cancer.