基于多元统计分析方法的心悸阴虚火旺证判别模型建立

目的：采用多元统计分析方法，建立心悸阴虚火旺证量化标准。方法：采用X2检验或秩和检验筛选出对阴虚火旺证与非阴虚火旺证有鉴别意义的指标，并利用SPSS17．0统计软件对其进行非条件的Logistic回归分析。结果：五心烦热、盗汗、口干或口苦、脉数、眩晕所建立的模型为最佳诊断模型，经检验阳性预测的准确率为82．7％、阴性预测的准确率为83．3％，符合统计学的判别要求。结论：建立的心悸阴虚火旺证判别模型，可为心悸阴虚火旺证证候客观化研究提供依据。

Establishment of discriminant model of palpitation with hyperactivity of fire due to Yin deficiency syndrome based on multivariate statistical analysis

Objective：To establish quantitative criteria of palpitation with hyperactivity of fire due to Yin deficiency syndrome（HFYDS） based on muhivariate statistical analysis. Methods：Picked out the variables that could differentiate HFYDS from non-HFYDS with chi-square test or rank sum test. The data were analysecl by Non-conditional logistic regression of SPSS 17.0. Results：The model including five center heat,night sweats,dry mouth or bitter taste in the mouth,rapid pulse,dizziness was the best diagnostic model. The positive predictive accuracy was 82.7% ,negative predictive accuracy rate of 83.3%,to provide an objective basis for the diagnosis of HFYDS. Conclusion：The re- search provide references for objectification of palpitations with HFYDS.