医学观察数据的元数据框架反标准化应用

目的：构建医学观察数据的语义模型和元数据框架,为医学观察数据的规范化表达和信息共享提供基础.方法：参照国际标准化组织ISO有关信息技术标准及SNOMED CT、LOINC等医学领域相关国际、国家标准,定义医学观察项目的属性和元数据规范,采用UML建模.结果：医学观察数据的语义模型将医学观察逐步分解,形成类的树状层次结构.第一层可分为体格检查、实验室检查等4个子类;类包含一组特定的属性;属性以数据元的形式,通过元数据进行描述;一组特定属性（数据元）的实例即为规范化的医学观察数据.结论：语义模型和元数据框架规范下的医学观察数据具有完整、清晰的语义和统一的格式.构建元数据框架有利于医学观察数据的标准化,可作为国家卫生数据字典的研发策略.

Development and Application of Metadata Framework for Medical Observation Data

Objective： To build semantic model and metadata framework for medical observation data with the purpose of standardized representation and sharing of such data. Methods： Referring to relevant ISO standards on information technology and other international and national standards/specifications in health domains such as SNOMED CT, LOINC, etc., we defined the model and metadata specifications for medical observation data using Unified Modeling Language （UML）. Results： The semantic model developed has a hierarchical structure by classifying super class medical observation, which is very general into more detailed specific observation items. The first level can be composed of four sub-classes as physical examination, laboratory tests, etc. Every class and subclass has a set of specific attributes, which are represented as data elements and can be described by a given set of metadata. Standardized medical observation data are instances of the attributes. Conclusions： The medical observation data conformed to such metadata specifications has comprehensive and unambiguous semantics with unified format. This metadata framework is beneficial to development of medical data standards at national level.