To explore the regularity of traditional Chinese patent medicines for the treatment of hyperlipidemia recorded in Newly Edited National Chinese Traditional Patent Medicines,the Composition Principles of Chinese Patent Drugs,New Drug Conversion Standard,the Compilation of National Standard for Traditional Chinese Medicines and Chinese Pharmacopoeia. Researchers extracted the information of prescriptions from these cases according to the inclusion and exclusion criteria. Then microsoft excel 2010 was used to conduct frequency statistics and count the frequency of traditional Chinese medicine. SPSS Clementine( ver.12. 0) and SPSS( ver. 18. 0)were adopted respectively for frequency analysis,association rules analysis,cluster analysis and factor analysis. Besides,KMO test and Bartlett spherical test were performed for factor adaptation test. Finally,a total of 173 traditional Chinese medicines were included,involving 94 Chinese patent medicine prescriptions. The frequency results of traditional Chinese medicine showed that there were 33 kinds of high-frequency traditional Chinese medicine,mainly including those for tonifying medicine,activating blood and resolving stasis and blood-stasis,and clearing damp. The association rules analysis found out 12 association rules of drug pairs,3-herb pairs of 25 and4-herb pairs of 6. Totally 11 medicine groups with relevance were respectively extracted by cluster analysis. KMO test and Bartlett spherical test indicated that the method was suitable for factor analysis and 11 common factors were respectively extracted by factor analysis. The association rules reflected the therapeutic method for tonify the liver and kidney,activating blood and resolving stasis. Cluster analysis and factor analysis showed the therapeutic method of Qi-enriching and Yin-nourishing,and the factor analysis focused more on removing blood stasis and dampness. The decision tree with hawthorn as the dependent variable reflects the importance of alisma orientalis and fructus schisandrae in the drug matching. In conclusion,data mining technique can comprehensively analyze the regularity of prescriptions of traditional Chinese patent medicine for hyperlipidemia,and is helpful for guiding the development of Chinese patent medicines and the clinical practice of traditional Chinese medicine.
Keywords: Chinese patent medicine; data mining; hyperlipidemia; medication law.