摘要
目的:应用分类树模型构建重症手足口病的预测模型,并评价其应用价值。方法:整群抽取河南省郑州市某医院2013年4月至6月住院治疗的221例发病时间≤72 h的手足口病患儿为研究对象,采用CHAID分类树算法建立重症手足口病的预测模型,采用错分概率Risk值、索引图及受试者工作特征曲线评价模型的应用价值。结果:所建立的分类树模型包括3层,共9个结点,共筛选出4个解释变量:精神差、易惊、热峰≥39℃、手足抖动;其中最为重要的预测因素为精神差和易惊。模型错分概率Risk值为0.045,模型拟合的效果较好。结论:分类树模型不仅能有效地拟合重症手足口病的风险预测,还可以对变量间的交互作用进行有效的筛选。
Aim： To establish a risk prediction model for severe hand-foot-mouth disease（ HFMD）,and to evaluate its application value for severe HFMD patients.Methods： A total of 221 cases of HFMD within 72 hours who admitted to a hospital in Zhengzhou from April to June of 2013 were cluster selected for questionnaire investigation.The clinical data and laboratory parameters of the patients were collected to analyze the main factors for severe HFMD by making the use of the CHAID classification tree algorithm.The value of the established model was evaluated by the Risk statistics,index map and ROC curve.Results： The model had 3 stratums and 9 nodes.There were 4 explanatory variables screened out in the model,including poor spirit,easy to panic,top temperature above 39 ℃ and shake of hands and feet.Poor spirit and easy to panic were the most important risk factors.The risk value of misclassification probability of the model was 0.045,and the classification tree model fitted the actuality very well.Conclusion： Classification tree model can not only properly predict severe HFMD in children,but also reveal the complex interaction effects among the factors.
出处
《郑州大学学报（医学版）》
CAS
北大核心
2015年第1期20-25,共6页
Journal of Zhengzhou University(Medical Sciences)
基金
国家自然科学基金资助项目81172740
关键词
重症手足口病
分类树
危险因素
预测模型
severe hand-foot-mouth disease
classification tree
risk factor
prediction model
