Leveraging Interoperable Electronic Health Record (EHR) Data for Distributed Analyses in Clinical Research: Technical...

Area of Science:

• Medical Informatics
• Clinical Research Informatics
• Health Data Science

Background:

• The Medical Informatics Initiative (MII) Germany established 38 data integration centers (DIC) across university hospitals to enhance healthcare and research using electronic health record (EHR) data.
• The HELP (Hospital-wide Electronic Medical Record Evaluated Computerized Decision Support System to Improve Outcomes of Patients with Staphylococcal Bloodstream Infection) study was a use case to demonstrate the value of these DIC.
• This clinical trial assessed the impact of a computerized decision support system for managing staphylococcal bacteremia.

Purpose of the Study:

• To present technical lessons learned during the implementation of an EHR data analysis infrastructure for a clinical use case.
• To outline challenges encountered and solutions developed during the initial implementation of the DIC infrastructure.
• To provide insights applicable to other research platforms utilizing EHR data, focusing on data definition, integration, and analysis.

Main Methods:

• An interdisciplinary team developed a data catalog and domain-specific information model for study evaluation.
• DICs created extract-transform-load (ETL) pipelines using Health Level Seven International (HL7) Fast Healthcare Interoperability Resources (FHIR) and MII core dataset profiles for data standardization.
• Analysis scripts were distributed for local data preprocessing, followed by central analysis of aggregated results.

Main Results:

• Significant heterogeneity in data quality and interoperability standards necessitated substantial harmonization efforts.
• Development of analysis scripts and data extraction required multiple iterative cycles and close collaboration with local data experts.
• The study demonstrated the feasibility of distributed EHR analyses, underscoring the need for data quality assessment, realistic planning, and multidisciplinary expertise.

Conclusions:

• Leveraging EHR data for clinical research presents challenges, particularly regarding data standards and harmonization efforts.
• Progress in digitization and interoperability frameworks offers potential for future improvements in EHR data utilization.
• Lessons learned can inform the development of standardized methodologies and infrastructures for sustainable EHR data integration in research.