As the foundation for its network, EHDEN adopted the Observational Medical Outcomes Partnership Common Data Model (OMOP CDM) [22-24]. The OMOP CDM is widely recognized for its “structure + content” approach whereby the tables and fields (structure) as well as the vocabulary (content) are standardized, allowing for integration of data across multiple systems while maintaining data integrity. The model also supports a wide range of data types, including electronic health records (EHRs), claims data, and patient registries.

The OMOP CDM is maintained by the Observational Health Data Science and Informatics (OHDSI) community, an open science effort that aims to improve health by empowering a community to collaboratively generate the evidence that promotes better health decisions and better care [25]. The open-source nature of OHDSI allows for continuous community-driven improvements, making it adaptable to emerging health care needs [26-28].

Any organization with access to a data source in Europe could apply to be included in the EHDEN network. In this context, a data source is defined as a distinct repository of health care-related data pertaining to a specific set of individuals. Except for the COVID-19 Rapid Collaboration Call, the 7 DP calls executed between September 2019 and October 2022 were aligned to similar timelines for DP identification, grant awarding, and initiation of data harmonization (Figure 1). In each call, candidate partner organizations with access to one or more electronic health care databases applied to the EHDEN Harmonization Fund for a grant to implement or enhance their database (Multimedia Appendix 1). DPs were selected based on 3 criteria: data impact (size, coverage, quality), network impact (track record, uniqueness within network), and readiness (willingness to participate, governance) (Multimedia Appendix 2), reviewed by a Data Source Prioritization Committee. Each application was reviewed and scored by 2 reviewers, and the top applicants per round were awarded a grant.

Once DPs were identified and grants awarded, each data source underwent standardization to the OMOP CDM. A crucial factor in EHDEN’s long-term sustainability and success was the recruitment and training of local small-to-medium enterprises (SMEs). These SMEs were brought on board through separate calls from the EHDEN consortium and certified via the EHDEN Academy education program concluded by an onsite or web-based training. SMEs played a pivotal role in supporting DPs throughout the extract, transform, and load (ETL) process by providing guidance and expertise. In total, 64 SMEs across 22 countries were certified by EHDEN to support DPs.

The ETL process followed by the DPs and supported by the SMEs was largely uniform, as outlined by Voss et al [24], and involved four key steps: (1) summarizing the native data, (2) creating the ETL specification, (3) mapping source vocabulary codes, and (4) implementing the ETL. This standardized approach ensured transparency in the followed procedure and adherence to the conventions in converting data sources to the CDM, while also allowing DPs to benefit from the SMEs’ specialized knowledge.

To promote semantic interoperability across heterogeneous health care systems, all source codes for diagnoses, medications, procedures, and measurements were mapped to standardized vocabularies (eg, SNOMED CT [Systematized Nomenclature of Medicine – Clinical Terms], RxNorm [Prescription Normalized Names], LOINC [Logical Observation Identifiers Names and Codes]) as required by the OMOP CDM. For example, the UK Biobank contributed data that included SNOMED CT-coded diagnoses from EHRs alongside custom-coded fields for self-reported conditions and blood pressure measurements. Mapping these nonstandard elements involved a combination of automated matching and manual curation followed by expert validation. This process facilitated consistent interpretation of clinical concepts across countries and enhanced the analytical interoperability of the network [29].

Payments were structured based on output; to receive full funding, DPs were required to meet 3 different milestones. The ETL specification document entitled DPs to 30%, ETL implementation and infrastructure released the next 40%, and the final 30% was received by the DP after final inspection of the harmonized data (Multimedia Appendix 3).

Each milestone was reviewed by an EHDEN consortium member who was part of the Milestone review committee. The ETL specification document required by milestone 1 was evaluated to ensure the mapping adhered to the OMOP CDM conventions and that the DPs or SMEs had a good understanding of the CDM and their own native data [23,30]. Milestone 2, the ETL implementation, had multiple review steps. The infrastructure was investigated to be sure the DPs were using a supported database platform [31]. The vocabulary mapping was evaluated to ensure most, if not all, source codes were included. The data quality dashboard (DQD) was developed by EHDEN Work Package 5 to provide a standard structure for quality assurance [32]. It was used by DPs throughout the ETL process to continually improve the standardized data sources [33]. As described by Voss et al [24], the median number of times a DP ran the DQD was 3 (IQR 2‐7). In general, conformance issues to the OMOP CDM were identified and addressed in initial runs, with more complex site-specific vocabulary mapping issues addressed in subsequent runs. DPs used the default failure thresholds in the first run. These were updated to reflect the nuances of each data source in later runs of the software. In milestone 3, final DQD results as well as the CDM Inspection Report were reviewed [34]. Once approved, the DP then entered their information into the EHDEN portal, an online platform open to the public designed to catalog metadata on each data source [35].

An individual data source was considered one entry in the EHDEN portal. Data sources were categorized based on country, person count, the levels of care represented (primary, secondary, or mixed), why a person was included, and how data were captured. These categories were ascertained from the data source description and metadata provided to the EHDEN portal and verified with the DPs.

There are 3 reasons persons could be included in an EHDEN data source (person inclusion), as defined by population, where a person enters the data source because they live in a certain geographical location or because they are registered with a practice or insurer; encounter, where a person enters the data source upon a visit to a health care provider for any medical reason; or disease, where a person enters the data source when satisfying specific criteria (ie, a person has a specific medical condition). The most restrictive reason was chosen as the classification for each data source.

We identified which types of data capture methods each source contained; it could be one or more of the following: EHR, a bill or adjudicated claim record for health services rendered (claim), measurements taken and results recorded (laboratory), a set of required information collected about participants in a registry (case report form), patient-reported data (survey), documents analyzed by pulling structured data from unstructured data using a natural language processing algorithm, or death information from an official source or government entity (death certificate). If the data source did not provide this information, they were categorized as unknown.

Patients or the public were not involved in the design, conduct, reporting, or dissemination plans of this research.

The varied health care data across Europe, as demonstrated by the summary of 210 data sources in EHDEN from 30 countries, underscores the critical need to generate evidence from more than one data source to comprehensively represent the health care needs or experiences of the entire European population. Across the person inclusion and care levels represented in the network, the data sources are well distributed, emphasizing how health care systems, populations, and data capture methods can differ substantially. While 74.7% (157/210) of the data sources report EHR as at least one of their data capture methods, only 40.4% (85/210) report EHR as their only data capture method. The other 34.3% (72/210) report some combination of EHR, laboratory, case report form, claim, natural language processing, and death register data, showcasing the tremendous heterogeneity of data available in Europe.

Prior initiatives like European Union–Adverse Drug Reactions (EU-ADR) and Innovative Medicines Initiative–European Medical Information Framework (IMI-EMIF) laid the groundwork for EHDEN, with learnings from those projects directly impacting this project [15,18,36,37]. EU-ADR demonstrated the feasibility of building a federated data network for large-scale drug safety monitoring in Europe using common data analysis files. IMI-EMIF made the first transition from using common input files like those in EU-ADR to the OMOP CDM, but it was not scalable due to the lack of funds and need for trained SMEs, both problems which EHDEN addressed.

The sustainability of the EHDEN initiative has been achieved through a combination of mechanisms that foster shared leadership, collaboration, and long-term value creation. One key factor has been the stimulation and enablement of both national and European collaborations. The establishment of OHDSI National Nodes has provided a platform for DPs within individual countries to collaborate, share best practices, and enhance data quality [38]. These nodes facilitate national-level harmonization while ensuring compliance with local regulations and coding systems, thereby strengthening the network’s integrity. Beyond this, EHDEN’s adoption in multiple European projects has further expanded its influence, including its pivotal role in enabling large-scale initiatives such as the Data Analysis and Real World Interrogation Network (DARWIN EU). This has also influenced how the European Federation of Pharmaceutical Industries and Associations (EFPIA) is standardizing its data, demonstrating EHDEN’s impact across sectors.

EHDEN has also delivered economic value by creating local ecosystems that support SMEs and DPs. Through the Harmonization Fund, EHDEN has injected resources into the European health care data landscape, with the return on investment yielding a multiplier effect. By recruiting and training SMEs through the EHDEN Academy, the initiative has built local expertise to support DPs throughout the ETL process, ensuring decentralized and sustainable support for the network.

One of the goals of EHDEN has been to standardize health data, akin to utilities like electricity or the internet, essential and accessible to a rapidly growing number of stakeholders across Europe. Now that EHDEN has transitioned from a project under IHI to the nonprofit EHDEN Foundation, the focus has shifted to sustaining, expanding, and improving the network while leveraging the harmonized data for evidence generation. This next phase aims to generate meaningful RWE for research and regulatory purposes. A recent report by The European Commission on the future of European competitiveness highlights EHDEN’s foundational role in shaping the future of the European Health Data Space, further solidifying its legacy as a critical driver of innovation and collaboration in European health care [39].

The success of EHDEN in harmonizing data to the OMOP CDM has led to significant advances in methods research and evidence generation [40-43]. Many of the DPs involved in EHDEN have used their standardized data to conduct analyses across a broad spectrum of use cases. For instance, several studies have been conducted to describe the natural history of diseases, the safety and effectiveness of treatments, and health care utilization patterns across diverse populations [44-46]. One clear example is a multinational network cohort study by Li et al [45], which used evidence generated from EHDEN DPs to characterize background incidence rates of adverse events of special interest related to COVID-19 vaccines. EHDEN’s standardized data has also enabled and improved the development of predictive models, allowing for personalized predictions of treatment outcomes and disease progression [47,48]. The harmonization of data has facilitated large-scale population-level studies, which are crucial for understanding trends in public health and informing health care policy decisions [49,50]. These examples of evidence generation illustrate the broad applicability of the data in the network, which serves both academic researchers and regulatory agencies. A regularly updated list of EHDEN-supported studies and publications is maintained on the EHDEN website, providing a comprehensive overview of the diverse applications of the network in regulatory, clinical, and methodological research.

A notable demonstration of EHDEN’s success is the network’s use in providing timely information on medicines under surveillance due to shortages in multiple European countries. More than 50 DPs contributed data to a study titled “Incidence, Prevalence, and Characterization of Medicines with Suggested Drug Shortages in Europe” [51]. This study represents the largest observational database study conducted across Europe, both in terms of the number of databases involved and its geographic scope. The findings will support European efforts to monitor the use of critical medicines, contributing to the global fight against medicine shortages.

The EHDEN data network has also affected other European collaboratives around RWD. IMI projects like PIONEER, BigData@Heart, EU-PEARL, and HARMONY also use the OMOP CDM and have partly continued the mapping work done in EHDEN [52-55]. In the EMA-commissioned DARWIN EU initiative, among the 20 DPs onboarded in the first 2 years, 16 are also EHDEN DPs [56].

Building on the progress achieved through the EHDEN network, several key areas offer opportunities for future development. One priority is fostering sustained engagement with DPs. Continuous collaboration will be essential to ensure that DPs remain active contributors to the network by regularly updating and improving their data contributions. Strategies to incentivize engagement, provide ongoing support, and ensure mutual value will be vital for the network’s long-term success, particularly as efforts shift toward more robust evidence generation and ongoing enhancements in data quality.

Expanding the network’s reach and optimizing its databases for specific research use cases are also key areas for growth. With 210 data sources currently included, there is a significant opportunity to onboard additional DPs and expand the network’s coverage across Europe. Future studies will also help identify gaps where further data optimization is required, such as refining mappings or addressing specific quality issues to ensure that the evidence generated is robust, reproducible, and generalizable.

Finally, the newly established EHDEN Foundation will play a critical role in these efforts. By securing funding and fostering collaborations, the Foundation can drive the onboarding of new DPs, address emerging research questions, and ensure that EHDEN continues to adapt to the evolving health care landscape. It also serves as a point of entry for external researchers, who may engage with the network and propose studies through its federated framework. These directions will position the network to remain a cornerstone for RWE generation in Europe, supporting both research and regulatory innovation.

The results of this study demonstrate that the identification, harmonization, and standardization of data sources through EHDEN have contributed significantly to understanding the diverse RWD landscape and advancement of evidence generation across Europe. These efforts are not only improving observational health research but are also influencing broader regulatory initiatives, such as DARWIN EU, which builds on the foundational work of EHDEN to leverage RWD for regulatory decision-making. Now that the initiative has transitioned to the EHDEN Foundation, there is an opportunity to focus even more on generating high-quality evidence, further solidifying the role of real-world data in improving health care and informing policy decisions across Europe.