Standardization Of Electronic Health Record Data

At present, various kinds of health systems constructed in healthcare sector of Chinahave not yet been interoperable, i.e., data collectedcan not be shared cross systems andregions due to the lack of unified standards, although they can collect and transmit datavertically. As a result, a regional information sharing platform, based on electronic healthrecords (EHR)as core need to be established. Through the platform, the requirement ofvarious of relevant institutions and personnel should be met, when information integritywith logical complete will be integrated with health data scattered in various organizations,and then the problems of "information island" and redundant construction of verticalsystems, which are troubling health informationizing construction for long, will beresolved. Therefore, setting up and improving the health information standard system fromthe national level will provide well-defined semantics and normative expressive patternfor health information from multiple sources, make the information shared across systems,and improve data availability, consistency, comparability, and usability. Data standard is an important part of health information standard system. Method ofdata standardization is mainly based on the data items in current health area of China. Themethod, extracting data items from the data record forms is simple, directive, and efficient;however, it may lead to endless data items and complicated relationships between the dataitems, while by no means torepresent them clearly resulting from the diversity andincreasing of data requirement of various specialties. As a result, data standardizationshould not only focus on the specification of data items one by one from the underlying,but also need to creat conceptual data model in the general background. Common dataelements will be extracted through the establishment of the class hierarchy, defining theproperty of classes, and standardize data types; that is standardized data definition andrepresentation methods, independent of specific applications, will be formulated with theconstraints of the model framework. And meanwhile, metadata standards should beestablished and a detailed description be made about the data element and its context, therelationship between the data element, and etc. Data element standard development shouldbe developed under the unified metadata framework, thus it is advantageous to thestandard orderly management, dynamic maintenance and effective utilization.The study is to manage data collection in EHR as a beginning, develope conceptualdata model in the health area, define the class property according to the object classes,standardize data types and property value, and make the specific data items in the healthrecords as the application examples of standardization expression of property. The methodof data element standardization, guided by conceptual data model, will provide thefoundation and basis for further normalizing the data items of current EHR and improvinghealth information data element from national level in the future. This study adopted thestandard descriptionproposed in ISO/IEC11179to formulate comprehensively thedefinition and presentation of data,providing the feasibility of the data management.The results of the study include:(1) The study established conceptual data model of health information and describedeight objection classes involved person, health organization, pharmaceutical, equipment and materials,medical observation, health expenditure, planning and intervention, anddiagnosis and evaluation, respectively, in the basis of that managing and analyzing1212data items, and abstracting object class described by data. Parts of the classes weresubdivided and formed hierarchical structure according to the health informationcharacteristic.(2) In the basis of Construction of a class hierarchy, property of various classes wereset up; data type was defined abiding by regulation of HL7for property according tocharacteristics of property; and the values of every property were made certain, that isascertaining the definition and expression of data element. Value set originated from oneor more decoding systems, such as international standard (like ICD-9,ICD-10, LOINC,SNOMED-CT), national standard and industrial standard (like data element domain codestandard),218standardized data element were formed. According to the type of every dataitems of EHR, which is defined as an example of standardized data element.(3) Referring to the basic property of the data elements proposed in registeredelement model and basic property of ISO/IEC1179-3, choosing21required property fromthem, and take health service object, marital status, personal marital status, marital statuscode, and personal marital status code as examples, describing in formulation through5metadata items, involving object class, property, data element concept, value domain, anddata element, respectively. And expressive examples of standardizing were provided.The conclusions of the study include:(1) Data standardizing technical route adopted in this study can be used as animportant reference for establishing and perfecting methodology system of the electronichealth record data standardizing, in China. First, conceptual data model of healthinformation was established beginning from managing data collection in EHR bybottom-up method. The data items have covered most information of public health andjunior health care territory, of which the information model established on the basis, isguaranteed the integrity of the model contents. Second, the study, through top-downtechnical route, has defined the class property according to object classes of model set; standardize data types and property value, and abstract standardized data element. Themethod has a clear and systemic theoretical basis and abundant international experience,be scientific and feasible.(2) The studyguided by conceptual data model framework and according to the classof model setting, define the class property according to the object classes, standardize datatypes and data element standard,meanwhile, thousands of data items of EHR were putunder the corresponding property according to their semantics respectively. The result,made as a local applicant example of standardizing data element, not only showed therationality of standard route mentioned above, but also proved the feasibility offormulating the current data items of health files with the method. As a result, the directionis indicated for current data standardization, and further standardization of the data itemswill obtain a benefit in future.(3) This study adopted the standard descriptionproposed in ISO/IEC11179toformulate meta data items such as object class, property, value domain, data elementconcept etc. It will make the expression of data semantics achieve higher and larger-scopeconformity and transparency. At the same time, abiding by metadata registration standard,it will establish the foundation for data standard management and dynamic maintenance innational level, as well as continued evolution of standard system.