Towards FAIR Clinical Registries: A Privacy-by-Design Framework for Migrating Legacy Document-Based Clinical Reports...

Area of Science:

• Medical Informatics
• Clinical Research Data Management
• Health Data Interoperability

Background:

• Clinical registries are crucial for research but often rely on unstructured legacy documents (Word, PDF).
• Manual data entry from these documents is time-consuming, error-prone, and hinders data reuse.
• Privacy regulations and network restrictions complicate data migration to modern research systems.

Purpose of the Study:

• To develop a privacy-by-design pipeline for automatic conversion of right-heart-catheterization (RHC) reports into structured, research-ready data.
• To improve data interoperability and facilitate scalable research infrastructure.
• To reduce manual effort in data entry and ensure compliance with privacy regulations like GDPR.

Main Methods:

• A privacy-by-design pipeline was developed using deterministic parsing and domain-specific mapping.
• The workflow enforces longitudinal consistency for repeated patient examinations.
• A mandatory anonymisation checkpoint is integrated before data transfer from the clinical domain.

Main Results:

• The pipeline successfully converts document-based RHC reports into structured data within the REDCap system.
• Deployed in a pulmonary hypertension registry, the system handles over 5000 patients.
• Auditable, reproducible outputs are generated, significantly reducing manual workload and ensuring GDPR compliance.

Conclusions:

• The developed pipeline offers a practical solution for modernizing document-based clinical registries.
• This approach advances FAIR-aligned (Findable, Accessible, Interoperable, Reusable) and scalable research infrastructures.
• Automated data conversion enhances research efficiency and data integrity while upholding patient privacy.