Common data element (CDE) and AI-Enabled approaches for clinical data Management: Workshop proceedings from the NIH...

Area of Science:

• Clinical data management
• Down syndrome (DS) research
• Health informatics

Background:

• Multisite, longitudinal, and multisystem clinical data present significant management challenges in Down syndrome (DS) research.
• The INCLUDE Project convened a workshop to address these challenges in data management.

Purpose of the Study:

• To evaluate challenges in managing complex clinical data for Down syndrome (DS) research.
• To identify approaches for improving data standardization and reuse in DS research.
• To align and modernize clinical data management (CDM) practices.

Main Methods:

• NIH workshop summarizing outcomes and lessons learned.
• Utilized common data elements (CDEs) and artificial intelligence (AI)-enabled tools for data management modernization.
• Focused on strategies for cross-study harmonization and interoperability.

Main Results:

• Consensus on foundational role of shared CDEs, ontologies, and data models for harmonization and interoperability.
• Identified electronic health record strategies for scalable data extraction and analysis.
• Recognized AI-enabled methods, including NLP with human review, can accelerate information abstraction and support harmonization.
• Prioritized governance for privacy, transparency, bias mitigation, and oversight in AI application to health data.

Conclusions:

• Combining standardized CDE-driven design with governed AI workflows reduces manual burden and improves data quality.
• Enables integrated multimodal research for Down syndrome (DS) and other complex clinical programs.
• Provides applicable lessons for the INCLUDE Project and broader clinical research initiatives.