血瘀证量化诊断流行病学调查数据的Logistic回归分析  

Logistic Regression Analysis of Epidemiological Survey Data on Quantitative Blood Stasis Syndrome Diagnosis

作　　者：姚魁武[1] 张勇 王阶[1] 武继涛[3] 方居正[3] 朱翠玲[3] 

机构地区：[1]中国中医科学院广安门医院,北京100055 [2]郑州市管城区疾病预防控制中心,郑州450004 [3]河南中医学院第一附属医院,郑州450008

出　　处：《世界科学技术-中医药现代化》2008年第4期30-33,共4页Modernization of Traditional Chinese Medicine and Materia Medica-World Science and Technology

基　　金：国家自然科学基金重大课题(90209011):基于计算智能的血瘀证量化诊断研究，负责人：王阶;国家973计划项目(2003CB517103):冠心病心绞痛病证结合的诊断标准与疗效评价体系研究，负责人：王阶。

摘　　要：目的:运用统计学方法发现对于血瘀证诊断具有较好区分能力的诊断项目。方法:在现代流行病学原则指导下进行血瘀证诊断的多中心、大样本临床资料收集,对其结果进行Logistic回归分析处理。结果:从中筛选出的协变量——性别、职业、齿龈色黑、肌肤甲错、眼周色黑、面色黑、五心烦热、腹部肿块、痛经、月经血块、舌质紫暗、舌体瘀斑、舌脉曲张、脉涩(共14个)的OR值大于1,提示为危险因素,它们有增加血瘀证诊断机会的危险性。对2004例患者依据回归方程进行回代分析,总预测正确率为92.9%。结论:Logistic回归分析可以筛选出对于鉴别血瘀证和非血瘀证诊断中对于因变量影响较大的指标。This is a study to explore the diagnostic items that are in a better position to identify the diagnostic blood stasis syndromes （BSS）. An epidemiological survey was made in three areas of China to collect the BSS data. A logistic regression analysis was made to analyze the database. 14 items were selected as major BSS risk factors. The total forecasting ratio of the retrospection test was 92.9% for the 2004 cases. It is believed that the logistic regression analysis method can be employed to screen the items that have a better contribution to the cause variables of both BSS and non-BSS.

关 键 词：血瘀证 流行病学调查 LOGISTIC回归分析 

分 类 号：R587.1[医药卫生—内分泌] R734.2[医药卫生—内科学]