Predictive Efficacy and Nomoplot Prediction Model for Kawasaki Disease Children Combined with Coronary Artery Lesion was Established Based on Logistic Regression
10.13241/j.cnki.pmb.2025.12.007
- VernacularTitle:基于Logistic回归建立川崎病患儿并发冠状动脉病变的诺莫图预测模型及其效能评价
- Author:
Chang ZHOU
1
;
Dong-yang LIU
1
;
Jun-xia WANG
1
Author Information
1. 西部战区总医院儿科 四川成都 610083
- Publication Type:Journal Article
- Keywords:
Kawasaki disease;
Coronary artery lesion;
Influencing factors;
Nomoplot;
Prediction model
- From:
Progress in Modern Biomedicine
2025;25(12):1969-1975
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To explore the establishing Nomoplot prediction model for Kawasaki disease(KD)combined with coronary artery lesion(CAL)children was established based on logistic regression,and to evaluate its predictive efficacy.Methods:The clinical data of 367 KD children who were admitted to our hospital from January 2014 to June 2024 were retrospectively analyzed,the children were divided into CAL(n=91)and NCAL(n=276)groups according to whether they had concurrent CAL.The influencing factors for CAL children combine with KD were analyzed by univariate and multivariate logistic regression.Nomoplot prediction model for CAL children combine with KD was constructed.The predictive efficacy of the Nomoplot prediction model for CAL children combine with KD was evaluated by receiver operating characteristic(ROC)curve.Results:The incidence of CAL in 367 KD children was 24.80％(91/367),univariate analysis results showed that,the proportion of delayed intravenous immunoglobulin(IVIG)treatment and the proportion of IVIG resistance in CAL group were higher than those in NCAL group(P＜0.05),the fever duration was longer than that in NCAL group(P＜0.05),and the age,hemoglobin(Hb)and platelet count(PLT)levels were lower than those in NCAL group(P＜0.05),erythrocyte sedimentation rate(ESR),B-type natriuretic peptide(BNP),C-reactive protein(CRP),procalcitonin(PCT),albumin(ALB)and alanine aminotransferase(ALT)levels were higher than those in NCAL group(P＜0.05).Delayed IVIG treatment,prolonged fever duration,elevated CRP,ESR and BNP levels were independent risk factors for CAL children combine with KD(P＜0.05),Older age,elevated Hb level were protective factor(P＜0.05).The predicted curve of the Nomoplot prediction model was in good agreement with the ideal curve,the area under the curve(AUC)of the model for predicting CAL children combine with KD was 0.971.Conclusions:Delayed IVIG treatment,fever duration,age and Hb,CRP,ESR and BNP are independent Influencing factors for CAL children combine with KD,the Nomoplot prediction model constructed based on the above logistic regression results has a high predictive value for CAL children combined with KD.
