A reference data model of a metadata registry preserving semantics and representations of data elements

Area of Science:

• Biomedical Informatics
• Data Management
• Clinical Data Standards

Background:

• Integrating and analyzing long-term clinical data from multiple sources presents significant challenges.
• Existing data warehouses and metadata registries often fall short in handling complex data dependencies.
• Current metadata registries primarily focus on data element descriptions, neglecting hierarchical and temporal relationships.

Purpose of the Study:

• To address the limitations of current metadata registries in managing complex clinical data.
• To develop and validate a reference data model capable of handling heterogeneous data definitions and representations.
• To enable revision and branching control for conceptually similar data elements over time.

Main Methods:

• Developed a reference data model adhering to ISO/IEC 11179 standards.
• Incorporated mechanisms for revision and branching control of data elements.
• Validated the model's ability to map hierarchical and temporal dependencies.

Main Results:

• The proposed reference data model successfully accommodates conceptually similar data elements with varying definitions and representations.
• The model allows for effective revision and branching control, crucial for longitudinal data analysis.
• Demonstrated improved integration and analysis capabilities for complex clinical datasets.

Conclusions:

• The ISO/IEC 11179-based reference data model provides a robust solution for managing heterogeneous clinical data.
• This approach enhances the integration and analysis of long-term clinical data by addressing hierarchical and temporal complexities.
• The validated model offers a significant advancement in clinical data management and standardization.