基于数据挖掘的消化系统癌性疼痛治疗方剂组方规律

Analysis of Prescription Patterns for Gastrointestinal Cancer Pain by Data Mining Techniques

目的运用数据挖掘技术探究消化系统癌性疼痛治疗方剂的用药规律。方法筛选并整理《肿瘤方剂大辞典》《肿瘤良方大全》《中华人民共和国药典》《国家中成药标准汇编》《中华人民共和国卫生部药品标准》等资料中所收录的消化系统癌性疼痛治疗方剂,运用Microsoft Excel 2019进行频次统计、IBM SPSS Modeler 14.1版软件进行关联规则分析、R4.0.2版软件中的arules以及arulesViz包对不同配伍原则下关联规则进行可视化、IBM SPSS Statistics 25版软件进行聚类分析以及因子分析。结果共纳入消化系统癌性疼痛治疗方剂123个,涉及中药304种。使用频次≥10的中药共33味,功效以活血化瘀、补虚、理气为主;关联规则分析结果显示共得到药对关联规则26条,以陈皮→木香药对的使用频次最高,3味中药药组关联规则14条,以乳香+没药→冰片的中药组合应用最多,关联规则可视化网络提示乳香-没药药对关联性最强;聚类分析得到10个中药组合;因子分析共提取出12个公因子,载荷系数最高中药的依次为莪术、乳香、三棱、没药。结论古今医方治疗消化系统癌痛呈现以行气止痛、活血化瘀为主,扶正补虚、抗癌解毒、化痰祛湿为辅的组方规律,与癌性疼痛的核心病机相符。

Objective To investigate the medication rules of cancer pain prescriptions in the digestive system by data mining techniques.Methods The formulas of cancer and pain in the digestive system were collected from the Dictionary of Tumor Formulas,The Complete Compilation of Good Tumor Formulas,The Chinese Pharmacopoeia,The National Compendium of Proprietary Chinese Medicines,and The Pharmaceutical Standards of the Ministry of Health of the People's Republic of China.The formulas were screened and organized.The analysis was performed by Microsoft Excel 2019 for frequency statistics,IBM SPSS Modeler 14.1 for correlation rule analysis and IBM SPSS Statistics 25 for cluster analysis and factor analysis.The statistic packages of arules and arulesViz from R4.0.2 were used to visualize association rules under different compatibility principles.Results A total of 123 digestive system cancer pain formulas involving 304 Traditional Chinese Medicine(TCM)were included in the study.Thirty-three TCMs were used more than 10 times,with the effects of promoting blood circulation and resolving blood stasis,tonifying deficiency,and regulating qi.The association rule analysis showed that a total of 26 rules were obtained,with citri reticulatae pericarpium→aucklandiae radix being the most frequently used,and 14 rules for 3 TCM drug groups,with olibanum+myrrh→borneol being the most frequently used.Visual network of association rule showed that the herb pair with the strongest relevance is olibanum-myrrh.A total of 10 TCM combinations were extracted from the cluster analysis.A total of 12 metric factors were extracted from the factor analysis,and the highest load factors were Curcuma longa,Boswellia serrata,Sanliang and Myrrh.Conclusion In the treatment of cancer pain of digestive system,the ancient and modern medical prescriptions mainly used the strategies of Qi circulation to relieve pain,blood circulation to remove blood stasis,supplemented by strengthening the body and tonifying deficiency,anti-cancer and detoxification,phlegm and dampness,which was consistent with the core concepts of pathogenesis of cancer pain in traditional Chinese medicine theory.