急性白血病患儿营养状况分析及营养不良风险预测可视化模型的构建与验证  

Analysis on nutritional status of children with acute leukemia and construction and validation of a visual model for predicting malnutrition risk

作　　者：鲍群谢 蔡建萍[1] 周倩[1] 刘文娟[1] BAO Qun-xie;CAI Jian-ping;ZHOU Qian;LIU Wen-juan(Department of Hematology,Children's Hospital Affiliated to Zhejiang University School of Medicine,Hangzhou,Zhejiang 310000,China)

机构地区：[1]浙江大学医学院附属儿童医院血液科,浙江杭州310000

出　　处：《中国妇幼保健》2024年第23期4670-4674,共5页Maternal and Child Health Care of China

基　　金：浙江省教育厅一般科研项目(Y202146530)。

摘　　要：目的分析急性白血病患儿营养状况,构建并验证急性白血病患儿营养不良风险预测可视化模型。方法选取2021年9月—2023年9月于浙江大学医学院附属儿童医院血液内科住院治疗的90例急性白血病患儿,于第1治疗周期结束次日采用患者主观整体评估法(PG-SGA)评估患儿营养状况。依据评估结果将90例急性白血病患儿分为营养不良组、对照组,采用单因素、多因素回归分析法明确急性白血病患儿营养不良的危险因素。采用R软件基于上述危险因素构建急性白血病患儿营养不良风险列线图模型,应用拟合优度检验、受试者工作特征(ROC)曲线评价列线图模型预测效能。结果经临床评估,90例急性白血病患儿中,中度营养不良32例,重度营养不良18例,营养不良发生率为55.56%。与对照组比较,营养不良组年龄<6岁、髓细胞白血病、化疗周期≥2个月、胃肠道反应较重、工具型喂养、主要照护者文化程度为高中及以下比例均较高(均P<0.05)。多因素logistic回归分析结果显示:年龄<6岁(OR=1.842,95%CI:1.607~2.152)、髓细胞白血病(OR=1.798,95%CI:1.516~1.994)、化疗周期≥2个月(OR=1.834,95%CI:1.663~2.225)、胃肠道反应较重(OR=1.845,95%CI:1.602~2.258)、工具型喂养(OR=1.805,95%CI:1.624~2.057)、主要照护者文化程度为高中及以下(OR=1.818,95%CI:1.658~2.274)是急性白血病患儿营养不良的危险因素。基于年龄、疾病类型、化疗周期、胃肠道反应、患儿喂养方式、主要照护者文化程度构建急性白血病患儿营养不良风险列线图模型,校准度验证结果显示实测值与预测值结果基本一致,Bootstrap内部验证法结果显示C-index指数为0.875(95%CI:0.735~0.929),证实该列线图模型校准度、区分度及预测效能均较好。结论急性白血病患儿住院期间营养不良发生风险较高,年龄、疾病类型、化疗周期、胃肠道反应、患儿喂养方式、主要照护者文化程度�Objective To analyze the nutritional status of children with acute leukemia(AL),construct and verify the visual model of malnutrition risk prediction in AL children.Methods A total of 90 children with AL hospitalized in Department of Hematology of Children's Hospital Affiliated to Zhejiang University School of Medicine from September 2021 to September 2023 were selected.The nutritional status of the children was assessed by patient subjective global assessment(PG-SGA)on the day after the end of the first treatment cycle.According to the clinical evaluation results,90 children with AL were divided into malnutrition group and control group.Univariate and multivariate regression analysis were used to identify the risk factors affecting malnutrition in children with AL.R software was used to construct a risk nomogram model of malnutrition in children with AL based on the above risk factors.Goodness of fit test and receiver operating characteristic(ROC)curve were used to evaluate the predictive efficiency of the nomogram model.Results Clinical evaluation showed that 32 cases of moderate malnutrition and 18 cases of severe malnutrition in 90 children with AL,with an incidence rate of malnutrition of 55.56%(50/90).Compared with control group,the proportions of cases less than 6 years old,myeloid leukemia,chemotherapy cycle≥2 months,severe gastrointestinal reaction,tool feeding,and main caregiver with high school education or below in malnutrition group were higher(P<0.05).Multivariate logistic regression analysis showed that less than 6 years old(OR=1.842,95%CI:1.607-2.152),myeloid leukemia(OR=1.798,95%CI:1.516-1.994),chemotherapy cycle≥2 months(OR=1.834,95%CI:1.663-2.225),severe gastrointestinal reactions(OR=1.845,95%CI:1.602-2.258),tool feeding(OR=1.805,95%CI:1.624-2.057),main caregiver education level of high school or lower(OR=1.818,95%CI:1.658-2.274)were risk factors for malnutrition in children with AL.Based on age,disease type,chemotherapy cycle,gastrointestinal reaction,feeding pattern,and main caregiver edu

关 键 词：急性白血病 儿童 营养不良 风险预测 

分 类 号：R733.7[医药卫生—肿瘤]