Mining And Analysis Of Drug Sampling Data In 11 Provinces Including Beijing, Zhejiang, Shanghai, Guangdong, Etc. From 2014 To The First Half Of 201

Drug Sampling and Testing is one of the most important method of supervision and management of post-marketing drugs.Drug regulatory department could detect substandard drugs then take appropriate executive action,crack down on illegal acts of manufacturing and selling substandard drugs,ensure the safety of the use of medication by public,via spot inspection for drug production,operation,usage of coherent organization.It has been nearly 60 years since establishing of drug inspection institution which for quality control,and gradually develop into an appropriate drug-testing-system according to the specific situation of China.In 2008,the year of starting drug sampling and testing plan in China which also contain provincial level,then accumulated huge amount of testing data because of these works.Facing such a huge drug testing data,it is a very important task to figure out how to carry on the reasonable and effective usage.Data mining technology is one of the methods,it can found the underlying rules of drug sampling and testing data,and guiding the work in the future for virtuous cycle.This thesis is based on the data in Drug Quality Announcement in some provinces and autonomous regions,researching on following works.1.Literature reviewThere are two parts of this section:1)Analysis and suggestions of sampling work for drug quality monitoring in China.2)The prospect research of data mining technology in the drug testing.Base on the literature review,to have a basic knowledge about currently situation and policy,work progress of drug sampling and testing in China.Meanwhile,by referring to the literature about data mining technology and comparing with it in other fields such as digging in adverse drug reaction(ADR)or clinical medication data,analyzing the possibility of data mining methods in drug sampling and testing.2.Overall situation of drug sampling and testingReferring to literature review and Internet investigation and research,this thesis determines Beijing,Zhejiang,Shanghai,Guangdong etc.as target provinces for data collection,the years is between 2014 to first half of 2016m,the total data is 27613 items which come from the website of Food and Drug Administration in provincial level.With scientific analysis technique and data mining methods,that we can find out problems in drug sampling and testing and potential trends in substandard drugs.The result indicates that drug quality is basically stable in above provinces,but there are large differences between them.Drug Quality Announcement released irregular in some provinces.There are large differences in distribution of dosage form,such as the proportion of tablet and injection seems much large.The substandard quality of Chinese medicinal materials is usually caused by sulfur dioxide overproof.3.Result analysis of drug sampling and testingIn order to find the potential relationship between drug substandard items and the property of drugs such as dosage form,type,route of administration and origin etc.,the weakness point of drug and the safety risk of quality,this thesis use Naive Bayes,association rules etc.to analysis the data in drug sampling and testing.4.Discussion and suggestions for the problem in drug sampling and testingThis thesis investigates the application of data mining technology in drug sampling and testing.Due to different calculation methods,the result shows different.Every method has its own benefits.Summary and analysis all the results,finding problems and propose suggestions,provide reference for drug sampling and testing.