浅议省级药品抽验合格数据的分析利用  

Brief Discussion on Analysis and Utilization of Qualified Test Data of Provincial-level Drug Sampling Inspection

作　　者：蔡丹宁 孙春艳 Cai Danning;Sun Chunyan(Hubei Institute for Drug Control,Wuhan 430075,China)

机构地区：[1]湖北省药品监督检验研究院,武汉430075

出　　处：《中国药师》2020年第9期1852-1854,共3页China Pharmacist

摘　　要：目的:初步探讨省级药品抽验合格数据分析利用的基本思路和手段。方法:通过各数据库的数据整合得到风险信号,在数据回顾分析的基础上应用数据挖掘软件的特征选择算法分析药品质量属性的影响因素。案例为复合维生素B片及其单方制剂。结果:上述品种2015年至今的检验数据表明,部分厂家复合维生素B片及其单方制剂质量不够稳定,批间差异较大,其维生素B1含量测定的主要影响因素为生产企业、有关物质含量和批号。结论:数据挖掘软件可以应用于随机性较强的数据并揭示数据中隐含的质量趋势,数据来源的扩充有利于深入认识药品风险规律。Objective:To discuss the basic ideas and means of utilizing qualified data of provincial-level drug sampling.Methods:The risk signal was obtained through data integration of each database.Based on the data review,the feature selection algorithm of data mining software was used to analyze the influencing factors of drug quality attribute.The multivitamin B and vitamin B1 tablets were used for the case study.Results:The inspection data of the above-mentioned varieties from 2015 to date suggested that the quality of multivitamin B and vitamin B1 tablets from certain manufacturer were not stable enough with a significant batch variance,and the main influencing factors for vitamin B1 assay were the manufacturers,the relevant substance content and batch number.Conclusion:Data mining software is potential in applying to the data with strong randomness and revealing the implied quality trends in the data.The expansion of data source will be conducive to the in-depth understanding of the related drug risk.

关 键 词：省级药品抽验 合格报告 数据挖掘 特征选择 

分 类 号：R97[医药卫生—药品]