Interoperability in clinical research: from metadata registries to semantically annotated CDISC ODM

Area of Science:

• Clinical informatics
• Data management in clinical research
• Health data standardization

Background:

• Clinical trial data capture using case report forms (CRFs) is often informal and lacks standardization.
• Existing metadata registries like NCI caDSR enable form creation using common data elements but lack standardized exchange formats.
• There is no current method for exchanging NCI-created forms into clinical trial management systems using the CDISC ODM standard.

Purpose of the Study:

• To develop a feasible mapping model between National Cancer Institute (NCI) forms and the Clinical Data Interchange Standards Consortium (CDISC) Operational Data Model (ODM).
• To analyze common data elements within NCI forms for frequency and uniqueness to inform the mapping process.
• To identify limitations encountered during the conversion and mapping of NCI forms to CDISC ODM.

Main Methods:

• Analysis of 3012 NCI forms to identify and evaluate common data elements.
• Development of a mapping model to translate NCI form structures and data elements into the CDISC ODM format.
• Identification of semantic and structural discrepancies between the NCI forms and CDISC ODM during the mapping process.

Main Results:

• A feasible mapping model was successfully created between NCI forms and CDISC ODM.
• Limitations were identified, including difficulties in mapping semantic codes from the caDSR registry to ODM items.
• Information such as the number of module repetitions was lost during the conversion process.

Conclusions:

• The developed mapping model demonstrates feasibility for converting NCI forms to CDISC ODM.
• Further refinement is needed to precisely specify the mapping of semantic concepts within CDISC ODM.
• Enhancing the precision of semantic mapping is crucial for accurate and complete data exchange in clinical trials.