Related Experiment Videos
Comparison of nutritional screening tools in pediatric patients undergoing hematopoietic stem cell transplantation
Lixian Han1, Yun Zhuang1, CaiQin Zhang1
1Department of Hematology, Children's Hospital of Soochow University, Suzhou, China.
Background:
Malnutrition is prevalent in children with hematologic malignancies undergoing allogeneic hematopoietic stem cell transplantation (HSCT) and is associated with elevated post-transplant infection risk. No standardized nutritional screening tool exists for this population. This study aimed to compare the performance of four tools (STRONGkids, PYMS, STAMP, and SCAN) in assessing malnutrition risk, and to explore the association between their risk classification and post-transplant infection.
Methods:
A total of 199 children with hematologic malignancies who underwent HSCT were enrolled. According to the Global Leadership Initiative on Malnutrition (GLIM) criteria, pre-transplant malnutrition was defined using a CT-derived skeletal muscle index (SMI) Z-score < -2 as the phenotypic criterion for reduced skeletal muscle mass, combined with the underlying hematologic malignancy as the etiologic criterion. This study compared the agreement between the risk classifications assessed by the four tools (STRONGkids, PYMS, STAMP, and SCAN) and GLIM-defined malnutrition diagnosis, as well as the correlation between these risk classifications and post-transplant infection.
Results:
Among the 199 children, 53 (26.6%) were diagnosed with malnutrition. The detection rates of high nutritional risk differed significantly among the four screening tools (PYMS 85.43%, SCAN 73.87%, STAMP 58.79%, STRONGkids 48.24%). In terms of screening performance, STRONGkids had an AUC of 0.801, a specificity of 67.8%, and the best agreement with the GLIM criteria (κ = 0.479); PYMS had the highest sensitivity (98.1%) but a specificity of only 19.2%. The 100-day post-transplant infection rate was 48.7%, and the high nutritional risk groups identified by all four tools had significantly higher infection rates (all P < 0.001). Among them, STRONGkids had an AUC of 0.743 and a Youden index of 0.487; the AUCs of SCAN, STAMP, and PYMS were 0.735, 0.681, and 0.642, respectively.
Conclusion:
STRONGkids showed potential application advantages in identifying malnutrition risk and discriminating post-transplant infection in HSCT children.