Construct a Prediction Model of Poor Prognosis of Anaphylactoid Purpura Children Based on Logistic Regression Analysis and Nomogram
10.13241/j.cnki.pmb.2025.12.010
- VernacularTitle:基于Logistic回归分析和列线图构建过敏性紫癜患儿预后不良的预测模型
- Author:
Xi LIN
1
;
Jing SUN
1
;
Ze-yu YU
1
;
Zhang-hua CHEN
1
;
Wei YANG
1
;
Xin-yu ZHANG
1
Author Information
1. 福州市罗源县医院儿科 福建 福州 350600
- Publication Type:Journal Article
- Keywords:
Henoch-schonlein purpura;
Children;
Prognosis;
Logistic regression analysis;
Nomogram;
Predictive model
- From:
Progress in Modern Biomedicine
2025;25(12):1989-1995
- CountryChina
- Language:Chinese
-
Abstract:
Objective:The purpose of this study is to explore the influencing factors of poor prognosis in henoch-schonlein purpura(HSP)children,and to construct a nomogram model and verify its validity through the analysis results of Logistic regression model.Methods:Collected the clinical data of 170 HSP children who were treated in Fuzhou Luoyuan County Hospital from January 2015 to May 2023 were retrospectively analyzed.They were divided into good prognosis group and poor prognosis group according to the prognosis after treatment.The clinical data and related biochemical indexes of different prognostic groups were compared.The factors of poor prognosis in HSP children were analyzed by Logistic regression model.A predictive model(nomogram)for poor prognosis in HSP children were constructed and evaluated its predictive performance.Prediction accuracy were evaluateed by receiver operating characteristic(ROC)curve.Results:170 HSP children,69 cases(40.59％)had poor prognosis and 101 cases(59.41％)had good prognosis.C-reactive protein(CRP),kidney injury at initial onset,platelet count(PLT),respiratory infection,recurrent rash,and immunoglobulin A(IgA)levels in poor prognosis group were higher than those in good prognosis group(P＜0.05).Multivariate logistic analysis showed that,recurrent rash,respiratory infection,elevated PLT,CRP and IgA levels were independent influencing factors of poor prognosis in HSP children(P＜0.05).The consistency index(C-index)of the column chart model established based on the results of multiple factor analysis was 0.798.The internal verification was carried out by Bootstrap self-sampling method,and the average absolute error of calibration curve was 0.017.ROC curve was drawn according to independent influencing factors and nomogram prediction probability,the area under the curve was 0.674(recurrent rash),0.649(respiratory infection),0.777(PLT),0.733(CRP),0.749(IgA)and 0.910(nomogram prediction probability)respectively.Conclusion:Recurrent rash,respiratory infection,PLT,CRP and IgA are independent factors affecting the prognosis of HSP children.The column chart prediction model constructed based on the above factors has certain predictive value for the poor prognosis in HSP children.
