Regional Electronic Health Records in Denmark: Research Potential for Clinical Epidemiology.

Denmark has a well-established tradition of utilizing national registries for epidemiological research. Healthcare services in Denmark are managed by five administrative regions and documented in electronic health records (EHRs). While certain EHR data are mandatorily transferred to national registries, the national registries capture only a fraction of the information available within regional EHR systems. In this review, we describe and discuss the context, content, and research potential of regional EHR data in Denmark for observational studies and pragmatic clinical trials. Throughout, examples and descriptions of data infrastructure primarily stem from one region (the Central Denmark Region). As each region governs its own healthcare delivery and IT systems, there are regional variations in EHR data sources and availability for research purposes. Overall, EHRs integrate clinical information, including structured entries (diagnoses, procedures, and treatments); semi-structured input of clinical information; free-text clinical notes; results from biochemical, microbiological, and pathological analyses; radiological images and descriptions; administered medicines; and administrative data such as appointment and procedure bookings. Regional EHRs include several data types unavailable or incompletely captured in national registries, including semi-structured and free-text clinical information, certain laboratory results, radiology descriptions, and hospital-administered or prescribed medicines before 2018. These data hold specific potential for research. Overall, the utilization of EHR data holds significant promise for improving both clinical trials and observational studies. In pragmatic trials, EHRs can facilitate patient recruitment and data collection on patient characteristics, outcomes, and adverse events within routine care, enabling larger, more generalizable study designs at lower costs. In observational research, EHR data can provide highly detailed clinical information and data types not captured in population-based registries. However, the use of regional EHR data also presents important challenges, including inter-regional heterogeneity in data systems and governance, variable data completeness, and methodological difficulties related to the extraction and analysis of unstructured clinical information. Despite these limitations, EHR data in Denmark provide large-scale, unique clinical information that complements population-based registries and represents an increasingly important resource for research. As data systems mature and the methods to utilize unstructured clinical information develop, the potential of these data to facilitate observational studies and pragmatic trials will continue to expand.