两种营养风险筛查工具用于危重患儿诊断效能研究

摘要

Objective To compare the ability to predict malnutrition in critically ill children between the STRONGkids and Screening Tool for the Assessment of Malnutrition in Paediatrics (STAMP).Methods A total of 647 critically ill children were assessed by the STRONGkids and STAMP.Results Using WHO child growth standards as gold standard (Z score),432 (66.8％) had nutritional risk,and 269 (41.6 ％) were diagnosed as malnutrition.The STRONGkids scored 16.8 ％ participants'low risk',60.3 ％'medium risk'and 22.9％'high risk',compared to STAMP categorisation:15.3％,39.1％ and 45.6％,respectively.Using medium and high malnutrition risk as cutoff value,the sensitivity,specificity,positive predictive value,negative predictive value and the area under the ROC curve of the STRONGkids were 92.4％,35.3％,74.2％,69.7％,and 0.741,and the corresponding values of the STAMP were 93.5％,33.0％,73.7％,71.7％,and 0.762.Conclusion Critically ill children are at high risk of malnutrition and malnutrition is common.Both the STRONGkids and STAMP have strong diagnostic value in identifying children at malnutrition risk for their high sensitivity and area under ROC curve,but the STAMP is more complex and time-consuming to use.%目的 探讨营养状况和生长风险筛查工具(STRONGkids)与儿科营养不良筛查工具(STAMP)用于危重患儿营养不良风险的诊断价值.方法 采用STRONGkids和STAMP对647例危重患儿进行营养风险筛查.结果 以WHO推荐的Z值为金标准,432例(66.8％)存在营养不良风险,269例(41.6％)诊断为营养不良.STRONGkids筛查结果显示营养不良低风险109例(16.8 ％)、中风险390例(60.3％)、高风险148例(22.9％).STAMP筛查结果显示营养不良低风险99例(15.3％)、中风险253例(39.1％)、高风险295例(45.6％).以中、高风险合并作为临界点时,STRONGkids和STAMP的灵敏度、特异度、阳性预测值、阴性预测值、ROC曲线下面积分别为92.4％、35.3％、74.2％、69.7％、0.741和93.5％、33.0％、73.7％、71.7％、0.762.结论 危重患儿营养不良风险和营养不良发生率高,两种筛查工具的灵敏度高和ROC曲线下面积较高,对营养不良风险有较高的诊断价值,但STAMP在实际操作中比STRONGkids更加复杂,耗时更多.