From Raw Data to FAIR Data: The FAIRification Workflow for Health Research

Area of Science:

• Health Informatics
• Data Science
• Digital Health

Background:

• The FAIR guiding principles (Findability, Accessibility, Interoperability, Reusability) promote data reuse.
• The GO FAIR initiative proposed a general FAIRification process, but it has limitations for specific data types like health data.
• Existing tools inadequately address the unique needs of health data management for FAIR principles.

Purpose of the Study:

• To design an open technological solution based on the GO FAIR FAIRification process, specifically addressing its gaps for health datasets.
• To develop a revised FAIRification workflow tailored to the technical, ethical, and legal requirements of health research data.
• To propose a technological architecture supporting the enhanced FAIRification of health data.

Main Methods:

• A common FAIRification workflow was developed by adapting existing steps and adding new ones for health data requirements.
• Analysis of technical barriers, ethical implications, and legal frameworks informed the workflow modifications.
• A technological architecture utilizing Health Level Seven International (HL7) FHIR resources was designed.

Main Results:

• A revised FAIRification workflow was created, addressing data curation, validation, de-identification, versioning, and indexing for health data.
• An open technological architecture was proposed, leveraging HL7 FHIR resources.
• The proposed solution supports the FAIRification of routine health care and research datasets.

Conclusions:

• The developed FAIRification workflow and HL7 FHIR-based architecture provide a framework for making health datasets FAIR.
• This approach enables the sharing and reuse of health research data within the research community.
• Researchers can benefit from this common framework to comply with increasing demands for FAIR health data.