张妮楠 曹馨宇 林睿凡 王斌 史华新 周洪伟 谢琪

摘 要：中醫常采用自然语言描述疾病症状，导致症状命名不统一，影响数据挖掘分析和临床疗效评价结果。该研究主要以中医临床专病诊疗数据为研究对象，示范性地探索和整理了临床诊疗数据中术语该如何规范化的问题，从语义分析的角度来规范化整理症状术语，在首选术语制定过程中引入术语属性和术语间关系的概念，解决了症状间多词一义、多义一词的问题。参照ICD-11中文版整理了疾病诊断术语;参照《中国药典》规范了中药药名;按照“治愈”“好转”“未愈”分类整理疗效评价术语。最终整理获得症状术语558条，其中首选术语164条，同义术语394条;诊断术语23条;疗效评价术语21条，规范后的数据可用于数据挖掘分析。

关键词：癫痫诊疗数据;数据规范化;首选术语;同义术语;术语属性

中图分类号：N04;R276文献标识码：ADOI：10.3969/j.issn.1673-8578.2021.02.006

Study on Standardization of Clinical Data of Epilepsy Diagnosis and Treatment//ZHANG Ninan， CAO Xinyu， LIU Ruifan， WANG Bin， SHI Huaxin， ZHOU Hongwei， XIE Qi

Abstract：Chinese medicine often uses natural language to describe the symptoms of the disease， which leads to inconsistent naming of the symptoms and affects the results of data mining analysis and clinical efficacy evaluation. Based on the diagnosis and treatment data of clinical special diseases of a famous Chinese medicine practitioner， this study explored on how to standardize the terminology in clinical diagnosis and treatment data. From the perspective of semantic analysis of standardization symptoms terms， we introduced the concept of the term properties and the relationships between terms in the process of establishing the preferred terms， and solved the problem of multi-word meaning and polysemy between symptoms. Also， we sorted out the disease diagnosis terms according to the ICD-11 Chinese version， standardized the names of traditional Chinese medicines according to Chinese Pharmacopoeia， and sorted the curative effect evaluation terms according to “cure”， “improvement”， and “unhealed”. We finally obtained 558 symptom terms， among which 164 were preferred terms， and obtained 394 synonymous terms， 23 diagnosis terms， 21 terms of efficacy evaluation. The standardized data can be used for data mining analysis.

Keywords： epilepsy diagnosis and treatment data; data standardization; preferred term; synonymous term; term properties

引言

开展中医真实世界研究，基于临床诊疗工作积累的病案，进行中医药个体化诊疗评价和分析，已经成为中医药专家学术经验传承、临床有效药物挖掘的重要途径。中医真实世界数据也已成为中医临床评价和药物监管决策证据体构成的重要来源。从临床诊疗病案中获得的真实世界数据，要通过数据相关性和可靠性评估其适用性，而数据规范化的工作更是数据研究与利用的首要关键环节，只有规范化的数据，才能够成为研究资料而发挥其应有的研究价值。

中医真实世界数……