基于Logistic回归分析中药注射剂不良反应预测研究.docVIP

基于Logistic回归分析中药注射剂不良反应预测研究 　　【摘要】目的：通过建立中药注射剂药物不良反应的Logistic回归模型，判断发生ADR时累及各组织器官的概率，为临床合理用药提供指导。方法：从广东省药品不良反应管理平台收集到阳江市2010至2015年间的有效中药注射剂不良反应报告1504份，对筛选出的数据进行清洗、属性规约和数据变换处理，运用SPSS Statistics 21统计分析工具，建立Logistic回归模型。结果：建立的Logistic回归模型总体预测正确率为633%，其中对皮肤及附件损害的预测正确率为846%，对全身性损害的预测正确率为836%，对其他组织器官系统的预测正确率均低于500%。结论：建立的Logistic回归模型能在一定程度上预测ADR时累及各组织器官的概率，该方法可为中药注射剂临床合理使用提供参考依据。 　　【关键词】中药注射剂；药物不良反应；Logistic回归 　　【中图分类号】R-3【文献标志码】 A【文章编号】1007-8517（2016）21-0115-05 　　Prediction of Adverse Drug Reaction of Traditional Chinese Medicine Injection Based on Logistic Regression Analysis 　　LIANG Ping1CHEN Yanfen2 　　1.Department of Pharmacy，Yangjiang Hospital of Traditional Chinese Medicine，Yangjiang 529500，China； 2.School of Traditional Chinese Medicine，Guangdong Pharmaceutical University，Guangzhou 510006，China 　　Abstract：Objective By establishing Logistic regression prediction model of Injection of traditional Chinese medicine，we can predict the probability of the occurrence of ADR in various tissues and organ to provide guidance for clinical rational use of drugs.Methods There are 1504 effective reports from 2010 to 2015 ，collected from the center for ADR monitoring of Guangdong.After cleaning， attribute reduction and data transformation， the data are obtained. With the Statistics SPSS 21 statistical analysis tools， a logistic regression model was established.Results The predict accuracy of Logistic regression prediction model of Injection of traditional Chinese medicine is 633% totality，in which damage to the skin and accessories of prediction accuracy is 846%， the systemic damage prediction accuracy is 836%， the accuracy of other tissues and organs system were lower than 500%.Conclusion The established Logistic regression prediction model of Injection oftraditional Chinese medicinecan predict the probability of the occurrence of ADR in various tissues and organto a certain extent.This method can provide reference for clinical rational use of Injection of traditional Chi