EMA Leverages Real-World Data Networks

How Is the EMA Using Real-World Data to Improve Regulatory Decisions?

The European Medicines Agency has been progressively building out the Data Analysis and Real World Interrogation Network, known as DARWIN EU, as a central pillar of its regulatory science strategy. The network connects electronic health record databases, claims data, and disease registries across multiple European countries, allowing regulators to commission studies that draw on routinely collected clinical information rather than relying solely on randomized trial data submitted by sponsors.

This shift toward real-world evidence is intended to fill knowledge gaps that emerge after a medicine is authorized. Randomized trials, while essential for establishing efficacy, often enroll narrower populations than those who eventually receive a drug in clinical practice. By querying anonymized records covering tens of millions of patients, the EMA can examine how medicines perform in older adults, pregnant women, people with multiple comorbidities, and other groups historically underrepresented in pivotal studies. The data also supports faster signal detection when unexpected adverse events emerge.

Why Does Animal Health Data Matter for Human Medicine?

The EMA's veterinary medicines work has gained prominence as antimicrobial resistance has become one of the most pressing global health threats. According to World Health Organization estimates, antimicrobial resistance is associated with millions of deaths globally each year, and inappropriate antibiotic use in livestock contributes to the emergence of resistant pathogens that can spread to humans through food, water, and direct contact.

By harmonizing data collection on veterinary antimicrobial sales and usage patterns across the European Union, regulators can identify high-risk practices and target interventions where they will have the greatest impact. The agency's One Health approach also enables faster response to zoonotic outbreaks, since shared surveillance infrastructure means that signals in animal populations can prompt early warnings for human health authorities and vice versa.

What Challenges Remain in Building a Pan-European Health Data System?

Despite progress, the EMA and its partners face substantial technical and governance challenges. Health data across European countries is recorded in different formats, languages, and clinical coding systems, making harmonization a major undertaking. The federated model used by DARWIN EU keeps data within each member state's jurisdiction and runs standardized analyses locally before aggregating only the results, which addresses many privacy concerns but adds complexity to study design.

The European Health Data Space regulation, which establishes a legal framework for secondary use of health data across the EU, is expected to provide clearer rules and expanded access in the coming years. Regulators have emphasized that public trust will depend on transparent governance, robust safeguards, and clear communication about how patient information is used to improve medicines and public health.