[Analysis and monitoring of traditional Chinese medicines for tumor based on data mining]

The traditional Chinese medicine(TCM) decoction pieces for treating tumors in China-Japan Friendship Hospital in both outpatient and inpatient departments from January 1 to December 31, 2018 were analyzed in this paper, and the statistical analysis on the frequency and proportion of TCM decoction pieces, as well as the average dosage and dosage range were conducted. Such data were then compared with Chinese Pharmacopoeia. At the same time, data mining association rules were used to study the compatibility of TCM in oncology, and finally, the drug use in TCM was discussed. The top 20(use frequency) TCM decoction pieces for tumors were mainly based on tonic medicines; the use frequency of toxic TCM decoction pieces was low, mainly of small poisonous pieces, with dosage exceeding pharmacopoeia. The drug combinations with higher frequency included Fried Atractylodis Macrocephalae Rhizoma-Poria Cocos(16.11%), and Astragali Radix-Poria Cocos(15.10%). Drug pairs with strong associations included Achyranthes Bidentata→Parasitic Loranthus, Coix Seed→Achyranthes Bidentata, Achyranthes Bidentata→Hairyvein Agrimony, Cuscutae Semen→Achyranthes Bidentata and so on. According to the use of drugs, the drug monitoring can be emphasized from the aspects of usage and dosage, selection of processed TCM, compatibility, decoction methods, and patient education. Pharmacists can analyze the characteristics and regularity of the use of TCM for tumors through data mining methods, and this can be a cutting point for drug monitoring.