摘要
目的基于关联规则和系统聚类分析探讨流行性感冒用药配伍规律。方法以“流行性感冒”为主题,在中国知网查找2009年1月—2019年1月文献,纳入符合标准的文献290篇,使用Excel 2013、SPSS Clementine 12.0、SPSS Statistics 23.0统计软件,将符合纳入标准方剂中的中药进行频数分析、关联规则分析和系统聚类分析。结果频数分析发现,在290例临床治疗文献中,单味中药甘草(204,70.34%)、连翘(168,57.93%)和金银花(151,52.07%)出现频率最高;在性味统计中,寒(16次,53.33%)、苦(19次,42.22%)最为常见;归经中肺经(25次,32.47%)、心经(14次,18.18%)最多;关联规则分析中得到关联强度最高的药对组合8组;聚5类为系统聚类分析中的核心组方。结论对临床治疗流行性感冒的有效临床处方运用现代信息技术进行分析,可以得到流行性感冒疾病最关键的药物配伍以及高发证型,对于减少临床大处方用药和开发新药均具有重要意义。
OBJECTIVE To explore the compatibility of influenza drugs based on association rules and cluster analysis.METHODS With“influenza”as the theme,from January 2009 to January 2019,290 eligible literatures were searched in CNKI.The frequency analysis,association rule analysis and systematic cluster analysis of Chinese medicines that met the criteria were carried out using Excel 2013,SPSS Clementine 12.0 and SPSS Statistics 23.0 statistical software.RESULTS Frequency analysis showed that among 290 clinical treatments,the frequencies of liquorice(204,70.34%),forsythia suspensa(168,57.93%)and honeysuckle(151,52.07%)were the highest,and cold(16 times,53.33%)was the highest in the statistics of sexual taste.Bitter(19 times,42.22%)was the most common;lung meridian(25 times,32.47%)and heart meridian(14 times,18.18%)were the most common;eight groups of drug pairs with the highest association intensity were obtained in association rule analysis;five groups were the core components in cluster analysis.CONCLUSION Modern information technology can be used to analyze the effective clinical prescriptions for influenza,and the most critical drug compatibility and high-incidence syndrome types can be obtained.It is of great significance to reduce the use of major clinical prescriptions and develop new drugs.
作者
马林纳
苗明三
MA Linna;MIAO Mingsan(Henan University of Traditional Chinese Medicine,Zhengzhou 450000,China)
出处
《中国现代应用药学》
CAS
CSCD
北大核心
2020年第7期837-841,共5页
Chinese Journal of Modern Applied Pharmacy
基金
国家国际合作基地(2016-65)
河南省产学研项目(182107000029)。
关键词
流行性感冒
中药
数据挖掘
influenza
traditional Chinese medicine
data mining
