Internal validation of two nutritional screening tools in hospitalized children and adolescents with cancer

In cancer patients, malnutrition is associated with poorer treatment outcomes and prolonged hospital stays, negatively affecting prognosis and survival. This study aimed to evaluate the internal validity of the nutritional screening tools Nutrition Screening Tool for Childhood Cancer (SCAN) and Screening Tool Risk on Nutritional Status and Growth (STRONGKids) in predicting nutritional risk in hospitalized pediatric cancer patients.

This was a retrospective observational study with pediatric cancer patients hospitalized between February and November 2019. ANPEDCancer was used as a reference to validate SCAN and STRONGKids due to its specificity in nutritional assessment for children and adolescents with cancer. Predictive values and agreement were analyzed. Patients were classified as at nutritional risk or not based on the tools. Nutritional status, anthropometric measures, biochemical markers, and length of stay (LOS) were compared between groups.

Of the 111 patients, 75.7 % (n = 84) were classified as at nutritional risk by SCAN, and 34.2 % (n = 38) as at high risk by STRONGKids. Patients at risk showed greater inadequacy in body composition measures and anthropometric indices. There was an association between nutritional risk as classified by the screening tools and the reduction of lean mass reserves (p = 0.039) and adipose tissue (p = 0.017). C-reactive protein was associated with nutritional risk only in STRONGKids, and with LOS only in SCAN. The agreement was 79.27 % for SCAN and 72.07 % for STRONGKids.

The SCAN demonstrated good precision in the nutritional screening of children with cancer, while STRONGKids was associated with inflamation.