目的：研究肝病患者的中医证型与检查指标的相关性，为肝病患者中医临床辨证提供客观依据。方法：收集2015年1月到2016年6月肝病住院患者的一般资料，以2015年肝病患者数据为分析组，2016年上半年患者数据作为测试组，运用逐步判别分析方法，筛选区分度高的指标，建立判别函数。结果：对2015年肝病患者的127个检查指标进行分析，筛选出13个指标，建立判别函数。利用判别函数对2015年224例患者数据进行回顾性预测，总的判别正确率为84.38%；并对2016年109例患者数据做前瞻性预测，总的判别正确率为74.31%。结论：逐步判别分析在肝病中医证型诊断中判别正确率较高，能为临床诊断提供参考。

This study was aimed to explore the correlation between traditional Chinese medicine (TCM) patterns ofhepatopathy and the objective inspection indicators so as to provide objective evidences for the syndrome differentiationand treatment of hepatopathy. General information of hepatopathy inpatients from January 2015 to June 2016 wascollected. In order to distinguish the most effective factors and establish the discriminant function, the data in 2015 wasselected as training samples with stepwise discriminatory analysis; and the data of the first half of 2016 was used astesting samples. A discriminant function was established by stepwise discriminant analysis, which including 13 inspectionindicators out of 127 indicators. Stepwise discriminant analysis was also employed to predict TCM syndrome patterns ofthe samples. The recognition accuracy of the training sample set 224 samples in 2015 was 84.38% and that of the testsample set 109 samples in 2016 was 74.31%. It was concluded that the stepwise discriminant analysis had a relativelyhigh discriminant accuracy rate. It can provide references in the clinical diagnosis.

老年病中药新产品湖北省协同创新中心项目（No.201506）：湖北省中医老年病数据资源管理平台构建研究，负责人：解丹；国家中医药管理局 中医临床研究基地业余建设科研专项课题（JDZX2012051）：中医治疗慢性乙型肝炎真实世界效果比较研究，负责人：李晓东。