王要强 吴瑞 管传涛 解中跃 程昆林 毕昌韦 章鑫 罗冬梅 李文喜

摘要 目的：遵循循证医学的原则，对201例肾病患者的中医药临床研究数据进行挖掘，发现慢性肾脏病的中医症状，为慢性肾脏病的诊断和治疗提供有效的参考。方法：运用频数分析法和关联规则分析对201例肾病患者临床数据进行分析。结果：患者的舌质主要是红和淡红，其频率分别为39.7%，33.3%;舌苔主要是白、白腻和薄白，其频率分别为34%，17.6%和15%;脉象主要是细，细滑，沉细脉象，其频率分别为20.5%，13.5%，10.3%。另外，患者的中医辨证主证主要是气阴两虚证和肝肾阴虚证，其频率分别达到46.3%，24.4%;中医兼证中湿热证、湿浊证和血瘀证占比较高，其频率分别为32.6%，29.4%与23%。且由关联规则分析可以得到中医主证和兼证与所对应的舌质、舌苔和脉象的关联性的强弱。结论：频数分析法和关联规则分析能够清楚地展示慢性肾脏病的中医征候规律，有助于慢性肾脏病的精准诊断和个性化治疗。

关键词 循证医学;慢性腎脏病;数据挖掘;频数分析;关联规则

Study on Traditional Chinese Medicine Syndromes of Chronic Kidney Disease Based on Data Mining Technique

WANG Yaoqiang，WU Rui，GUAN Chuantao，XIE Zhongyue，CHENG Kunlin，BI Changwei，ZHANG Xin，LUO Dongmei，LI Wenxi

（School of Mathematical Science and Engineering，Anhui University of Technology，Ma′anshan 243002，China）

Abstract Objective：To mine the data of 201 patients with kidney disease from TCM clinical research，and find the TCM symptoms of chronic kidney disease，which can provide an effective reference for the diagnosis and treatment of chronic kidney disease following the principle of evidence-based medicine.Methods：Frequency analysis and association rule analysis were used to analyze the clinical data of 201 patients with nephropathy.Results：The tongue of the patients was mainly red and light red，with frequencies of 39.7% and 33.3%，respectively.The tongue coating was mainly white，white greasy and thin white，with the frequencies of 34%，17.6% and 15% respectively.The pulse was mainly fine，smooth and sinuous，with frequencies of 20.5%，13.5% and 10.3%，respectively.In addition，the main syndromes of TCM syndrome differentiation in patients were qi and yin deficiency syndrome and liver and kidney yin deficiency syndrome，with the frequency of 46.3% and 24.4% respectively.Among the TCM syndromes，dampness and heat syndrome，dampness and turbidity syndrome and blood stasis syndrome were relatively high，accounting for 32.6%，29.4% and 23%，respectively.The correlation between the main syndrome and the concurrent syndrome and the corresponding tongue texture，tongue coating and pulse can be obtained by association rule analysis.Conclusion：Frequency analysis and association rule analysis can clearly demonstrate the Chinese medicine symptom rule of chronic kidney disease，which is helpful for accurate diagnosis and personalized treatment of chronic kidney disease.

Keywords Evidence-based medicine; Chronic kidney disease; Data mining; Frequency analysis; Association rules

中图分类号：R242文献标识码：Adoi：10.3969/j.issn.1673-7202.2020.24.029

循证医學崛起于20世纪90年代[1-2]，狭义上的循证医学为循证临床实践，强调医生在临床工作中应依据最佳的研究证据[3]，结合自己的临床经验，医疗条件及患者的情况与选择，制定合理的诊治方案。广义上的循证医学是在一切医疗卫生服务的循证实践中，强调制定宏观医疗卫生决策时遵循研究证据的重要性和必要性[4]。

循证医学不同于传统医学。传统医学是以经验医学为主，即根据非实验性的临床经验、临床资料和对疾病基础知识的理解来诊治患者。……