Introducing an Ontology-Driven Pipeline for the Identification of Common Data Elements.

Area of Science:

• Clinical Informatics
• Biomedical Data Science
• Health Services Research

Background:

• Common Data Elements (CDEs) are crucial for data sharing, comparability, and meta-analyses in clinical research.
• Manual CDE development is often laborious and time-intensive, hindering research progress.
• Standardization of data is essential for robust scientific inquiry and reproducible results.

Purpose of the Study:

• To introduce an automated pipeline for efficient identification, aggregation, and ranking of Common Data Elements (CDEs).
• To leverage clinicaltrials.gov (CTG) data and the Medical Subject Headings (MeSH) ontology for CDE generation.
• To reduce the burden of CDE development in clinical research domains.

Main Methods:

• An automated pipeline was developed to process study outcomes from clinicaltrials.gov.
• The Medical Subject Headings (MeSH) ontology was employed to group and rank candidate CDEs by disease.
• The pipeline was initially tested on an emerging research domain to assess its efficacy.

Main Results:

• The automated pipeline successfully identified, aggregated, and ranked relevant CDEs.
• The generated CDEs aligned with existing recommendations in the tested research area.
• The system demonstrated potential for significant time and resource savings in CDE development.

Conclusions:

• Automated CDE generation using structured data from clinicaltrials.gov and MeSH offers a promising solution.
• Further development is warranted to enhance automated methods for CDE creation.
• This approach can accelerate data harmonization and facilitate large-scale clinical research.