儿童胃肠积热自评诊断量表的研制和评价

Development and validation of self-rating diagnostic scale for gastrointestinal heat retention syndrome in children

目的:研制和评价儿童胃肠积热自评诊断量表。方法:基于既往研究确定儿童胃肠积热自评诊断量表的初步条目池,采用Lasso回归分析筛选量表的核心条目,将研究数据随机分为训练集和测试集,在训练集上采用二元Logistics回归模型构建诊断模型,将系数转化为整数,并根据Youden指数选择模型最佳阈值,在测试集上采用诊断性试验测评量表诊断性能。结果:自评诊断量表的初步条目池中共包含24个问诊条目,经Lasso分析筛选得到10个核心条目。二元Logistics回归诊断模型的条目按系数从高到低排列,依次为大便干结、食欲异常、小便色黄、夜卧不安、排便次数减少、口臭、手足心热、夜间汗出、烦躁易怒、大便臭秽,量表的界值为133分。自评诊断量表在测试集上,敏感度、特异度、准确率分别为90.14%、84.75%、88.56%。结论:形成了儿童胃肠积热自评诊断量表,其具有较高的诊断能力。

Objective:To develop and validate the self-rating diagnostic scale for gastrointestinal heat retention syndrome(GHRS)in children.Methods:The initial items of the self-rating diagnostic scale for GHRS was determined according to previous studies.The core items of the scale were screened by Lasso regression analysis.The data was divided into training set and testing set.The binary Logistic regression model was used to establish the model on the training data,and the coefficients were converted into integers.The optimal threshold was selected according to Youden index.Finally,diagnostic test was used to evaluate the diagnostic performance of the scale on the testing data.Results:The preliminary items of the self-rating diagnostic scale contain 24 items,and 10 core items are obtained through Lasso regression analysis.The items of diagnostic model were constructed by the binary Logistic regression.The items ranked by the coefficients from high to low were dry stool,abnormal appetite,yellow urine,restless sleep at night,reduced frequency of defecation,halitosis,feverish feeling in palms and soles,sweating at night,vexation and irascibility,smelly stool.The optimal threshold of the scale was 133 points.The sensitivity,specificity and accuracy of the scale on the testing set were 90.14%,84.75%and 88.56%respectively.Conclusion:The self-rating diagnostic scale for GHRS in children is developed,which has high sensitivity and specificity.