基于慢性阻塞性肺病(COPD)急性加重期文献的多元统计方法在证候研究的应用探讨

目的:基于COPD急性加重期的文献,在研究中医证候的分布中,探讨多元统计方法如因子分析、聚类分析等方法的应用.方法:检索了近10余年COPD急性加重期的文献资料,建立相应的数据库,采用SAS6.12统计软件包进行因子分析、聚类分析等.结果:因子分析提取出疾病层面的症状如咳嗽、咯痰和喘促及17个公因子;聚类分析得到3类证候属性--痰热壅肺、肺肾气虚、痰瘀阻肺,以及每类属性的特征.结论:因子分析和聚类分析在一定程度上能达到降维和消除共线性、由非线性向线性转化的目的,二者的联合使用有利于揭示证候的分布规律.

Investigate of Multivatiate Statistics Based Literature on Distribution of TCM Syndromes of COPD at the Acute Stage of Exabasedoidtion

Objective:to probe methodological application of multivatiate statistics such as factor analysis and cluster analysis,based literature on distribution of TCM syndromes of COPD at the acute stage of exabasedoidtion.Methods:literatures about COPD at the acute stage of exacerbation in near over 10 years were collected and databank was established.SAS 6.12 software was adopted factory analysis and cluster analysis.Results:it was extracted that symptoms such as cough,stethocatharsis and asthma and seventeen public factors by factory analysis.Syndromes attribute and characteristic were obtained by cluster analysis.Conclusion:combined use of two methods makes for revealing distribution regularity of TCM syndromes.