T/JSIA 0004-2024
Specification for healthcare big data quality management (English Version)
- Home
- T/JSIA 0004-2024
- Standard No.
- T/JSIA 0004-2024
- Language
- Chinese, Available in English version
- Release Date
- 2024
- Published By
- Group Standards of the People's Republic of China
- Latest
-
T/JSIA 0004-2024
- Scope
- 4.2 Basic Principles of Data Quality Management
Data quality management should adhere to the following principles:
a) Relevance: Refers to whether the collected statistical data is useful and meets user needs;
b) Accuracy: Data must be true, reflecting actual conditions with minimal errors;
c) Timeliness: Provide data promptly;
d) Comparability: International and domestic comparability; consistent statistical口径保持一致;annual interconnectivity;
e) Completeness: Content of data elements should be complete without any omissions.
f) Simplicity: Information should be concise, brief, and to the point;
g) Availability: It is convenient to obtain data and related statistical information consultation services;
h) Comprehensive coverage: Cover every stage of the lifecycle of data from planning, acquisition, storage, sharing, maintenance, application to disposal.
4.3 Process for Data Quality Management
Data collection, format checking, content validation, data integration, data review.
5 Grading System
5.1 Data Quality Grading
a) A grading standard for data quality should be established based on the completeness of reported data and user feedback. Results should be dynamically updated.
b) At least two levels of data quality: Level I and Level II.
Data Quality Level I: All data conforms to data standards and passes both format and content validation;
Data Quality Level II: All required data conform to data standards and pass format and content validation, but optional data do not fully comply with the standard requirements.
Data at Level II are primarily for subsequent corrections and updates.
5.2 Data Usage Grading
Based on the meaning, content, and scope of application, set a grade for the use of data, managing openness and sharing ranges according to grades. Typically, it is divided into online sharing and offline sharing:
a) Online Sharing: Users can query, browse, and download data via the Data Engineering Data Center Service Portal, dedicated data center service portals, co-built data center service portals, etc., including public sharing and authorized sharing.
Public Sharing refers to registered users downloading publicly available data after logging in.
Authorized Sharing involves partially private data being downloaded online with authorization based on the functions of each unit, granting corresponding privacy data download permissions.
b) Offline Sharing: Includes application-based sharing and extended sharing. Application-Based Sharing refers to users applying for data usage from relevant data centers, obtaining data through optical disc distribution after approval by the respective center's supervisory authority under safety technical support. Extended Sharing involves users requesting more specialized and in-depth processing of data. Data producers process the data according to user applications and provide it on optical discs with safety technical support.
5.3 Data Security Grading
a) Sensitivity Domain Identification
Certain collections of highly classified data are known as sensitivity domains. "Grading" involving sensitive data is not business-driven but determined by their inherent attributes, i.e., field meanings indicate the corresponding sensitivity domain. Commonly used methods include dividing sensitivity domains into public sensitivity domains (from a legal perspective), industry-specific sensitivity domains (from an industry standard perspective), and organizational sensitivity domains (from internal standards). Public Sensitivity Domains and Industry-Specific Sensitivity Domains are typically defined in regulations, while Organizational Sensitivity Domains require understanding of the business system by the personnel involved.
b) Determining Sensitivity Levels
Unlike categorizing data based on business drivers, sensitivity levels are determined by the level of secrecy. Methods for determining sensitivity levels can be based on the scope and impact range, objects, and severity caused by a potential data breach.
c) Categorizing Data Types
Typically, business domains are classified into "superclasses," "parent classes," "child classes," or more detailed subclasses, with data domains serving as the smallest classification.
d) Metadata Assigning to Sensitivity Domains
After identifying sensitivity domains, fields should be assigned to these domains for subsequent grading operations. If metadata management capabilities exist, or if fields have been pre-processed during the identification of sensitivity domains, this step can be skipped. Otherwise, fields need to be assigned to sensitivity domains, a process that may rely on intelligent discovery software.
e) Data Categorization and Consolidation
Data domains can be categorized via building a metadata management system. If such a system is not built, each field in every table of every database involved in the business systems needs to be categorized and graded manually. This process can also use intelligent software for assistance.
f) Multiple Classification and Grading Systems
In practical applications, due to varying business focus points, it may be necessary to establish multiple classification and grading systems to address different business demands. If compliance with regulatory requirements is needed, the primary requirement is to meet legal and regulatory standards. This work can be conducted by a legal and consulting team based on the scope of business as defined in relevant laws and regulations.
T/JSIA 0004-2024 history
- 2024 T/JSIA 0004-2024 Specification for healthcare big data quality management
- 2023 T/JSIA 0004-2023 Green Software Park Data Management Capacity Building Guide
- 2022 T/JSIA 0004-2022 Evaluation Specifications for Key Software Enterprises within the Planning Layout of Jiangsu Province
Topics on standards and specifications
Standard and Specification