Design of a Privacy-Preserving ETL Dataflow for Federated ICU Data Reuse in INDICATE

Area of Science:

• Health Informatics
• Data Privacy
• European Health Data Space (EHDS)

Background:

• Intensive care unit (ICU) data holds significant research potential.
• Secure and privacy-preserving data sharing is crucial for health research.
• The European Health Data Space (EHDS) aims to facilitate data access.

Purpose of the Study:

• To design a privacy-preserving dataflow for the INDICATE project.
• To enable the secure reuse of ICU data within the EHDS framework.
• To ensure data quality, interoperability, and regulatory compliance.

Main Methods:

• Development of a federated framework for secure data reuse.
• Implementation of a privacy-preserving Extract, Transform, Load (ETL) dataflow.
• Standardization and validation of ICU data using OMOP Common Data Model and HL7 FHIR.

Main Results:

• A robust ETL dataflow for local ICU data standardization and validation.
• Harmonization of ICU data ensuring semantic interoperability and data quality.
• A secure system where only aggregated study results are uploaded, protecting patient-level data.

Conclusions:

• The INDICATE project provides a secure and compliant method for ICU data reuse.
• The designed dataflow facilitates federated analysis while maintaining patient privacy.
• This framework supports the EHDS goals for enhanced health data research.