Keywords: electronic medical record, secondary data, health services research, semantic information

Background:

Secondary use of primary data from practice management systems for health services research is feasible; however, standardized formats, quality-assured comparability, interoperability between data storage systems, and open interfaces are missing.

Research questions:

To identify indispensable data fields (variables, parameters) suitable for secondary use across the Health Data Lab (FDZ), the Core Data Set of Medical Informatics Initiative (MII-KDS), and a generic RADAR data utilization framework.

Method:

Systematic literature review, analysis of source documentation. Data offerings from FDZ and MII-KDS were mapped to 11 semantic RADAR domains—diagnoses, medications, laboratory results, findings, therapies, other procedures, temporal data, patient demographics, practice-level data, payers, and billing information—for availability, granularity, and content alignment.

Results:

Semantic categorization reveals three essential domains:
Sociodemographic information: Basic attributes of patients are generally retrievable, socioeconomic status indicators are missing.
Biomedical: Diagnostic and medication data are consistently documented. Laboratory results, findings, and procedures appear as billing codes in FDZ dataset. MII-KDS allows for structured laboratory results representation.
Care pathways: FDZ encompasses longitudinal SHI-data from out- and inpatient care. MII-KDS records individual treatment episodes with procedural details, its Digital Progress Hubs aim to cross-sectoral longitudinal pathway-based data linkage.

Conclusions:

Neither FDZ nor MII-KDS alone sufficiently addresses all requirements for primary care research. FDZ billing data support population-based analyses restricted to SHI populations and lack key clinical data such as laboratory values or procedural details. MII-KDS provides detailed clinical documentation following international interoperability standards but does not yet capture cross-sectoral care pathways. All datasets lack socioeconomic variables.
Demographic baseline data, ICD-coded diagnoses, and medication information in FDZ and MII-KDS represent indispensable components for primary care research. Absence of laboratory values in FDZ significantly constrains epidemiological or clinical analyses to SHI patients. Effective utilization necessitates data integration from multiple sources: FDZ for population-based SHI care pathways, MII-KDS for structured individual clinical parameters, additional data linkage to incorporate socioeconomic information.

Points for discussion:

Which data fields (variables, parameters) are essential for research?

Are primary care EMR data "research-ready" in your country?

How to standardize European EMR information exchange for patient and research benefit?