川崎病患儿实验室检测指标分析及预测模型的建立  

Analysis of Laboratory Test Indexes and Establishment of Prediction Model for Children with Kawasaki Disease

作　　者：徐琨[1] 段炤[1] 邓宇虹[1] 徐晗 XU Kun;DUAN Zhao;DENG Yu-hong;XU Han(Department of Rheumatic Immunology Jiangxi Provincial Children’s Hospital,Nanchang 330006,China;Department of Pediatrics Jiangxi Provincial Children’s Hospital,Nanchang 330006,China)

机构地区：[1]江西省儿童医院风湿免疫科,南昌330006 [2]江西省儿童医院儿内科,南昌330006

出　　处：《实用临床医学（江西）》2022年第4期44-47,共4页Practical Clinical Medicine

基　　金：江西省卫生健康委员会科技计划项目(20203737)。

摘　　要：目的 分析川崎病患儿实验室检测指标的特点,并建立合适的预测模型。方法 收集2020年10月至2021年12月江西省儿童医院收治的川崎病患儿(n=59)和发热患儿(n=61)的临床信息,包括性别、年龄、白细胞计数(WBC)、谷丙转氨酶(ALT)、谷草转氨酶(AST)、谷氨酰转肽酶(GGT)、血尿素氮(BUN)、血肌酐(CR)、尿酸(UA)、C-反应蛋白(CRP)、血清淀粉样蛋白A(SAA)、血小板计数(PLT)、红细胞沉降率(ESR)、降钙素原(PCT)、血小板分布宽度(PDW)和血清补体C1q。采用ROC曲线分析评价各指标对于区分川崎病患儿和发热患儿的性能,并采用Logistic回归分析筛选出合适的指标建立预测的列线图模型。结果 与发热患儿相比,GGT、ALT、UA和WBC在川崎病患儿中含量显著升高,并表现出具有一定的特异性。WBC、AST、CRP、PLT和ESR 5种指标对于鉴别诊断川崎病患儿和发热患儿表现较好(AUC≥0.80),特别是ESR表现最佳(AUC=0.91,95%置信区间0.84~0.97)。由WBC、CRP和ESR建立的预测模型可较好地鉴别诊断川崎病患儿和发热患儿(AUC=0.93)。结论 实验室检测指标对于诊断川崎病有着一定的作用,由WBC、CRP和ESR建立的预测模型可应用于川崎病的临床诊断。Objective To analyze the characteristics of laboratory test indexes in children with Kawasaki disease,and to establish a suitable prediction model.Methods Clinical data of 59 children with Kawasaki disease and 61 children with fever admitted to Jiangxi Children’s Hospital from October 2020 to December 2021 were collected,including gender,age,white blood cell count(WBC),alanine aminotransferase(ALT),aspartate aminotransferase(AST),glutamyltranspeptidase(GGT),blood urea nitrogen(BUN),serum creatinine(CR),uric acid(UA),C-reactive protein(CRP),serum amyloid A(SAA),platelet count(PLT),erythrocyte sedimentation rate(ESR),procalcitonin(PCT),platelet distribution width(PDW),and serum complement C1 q.The ROC curve analysis was used to evaluate the performance of each index in distinguishing children with Kawasaki disease from those with fever,and logistic regression analysis was performed to screen out appropriate indexes to establish a predictive nomogram model.Results Compared with febrile children,the levels of GGT,ALT,UA and WBC significantly increased and showed a certain specificity in children with Kawasaki disease.WBC,AST,CRP,PLT and ESR exhibited good performance in differentiating Kawasaki disease from fever(AUC≥0.80),especially ESR(AUC=0.91,95% confidence interval 0.84-0.97).The prediction model established by WBC,CRP and ESR could distinguish the children with Kawasaki disease from those with fever(AUC=0.93).Conclusion The laboratory test indexes play a certain role in the diagnosis of Kawasaki disease,and the prediction model established by WBC,CRP and ESR can be applied to the clinical diagnosis of Kawasaki disease.

关 键 词：川崎病 实验室检测指标 列线图 

分 类 号：R725.9[医药卫生—儿科]