Follow us on X
Follow us on Linkedin
Real-world-data enabled assessment
for health regulatory decision-making

Halfway through: Two years of the REALM project

Two years ago, 15 partners from six countries kicked off the EU-funded REALM (Real-world data Enabled Assessment for Health Regulatory Decision-Making) project, which aims to revolutionise the evaluation and certification process of medical device software. The REALM platform is designed as a collaborative regulatory environment integrating real-world data (RWD) with advanced tools for assessing medical devices and healthcare innovations.

Coordinated by Prof. Michel Dumoniter (UM) and Dr Gökhan Ertaylan (VITO), the overarching goal of REALM is to build an inclusive, transparent, and standardised platform that enables health technology assessment agencies, regulatory bodies, healthcare professionals, and software developers to assess medical device software more effectively. By providing a decentralised, secure evaluation framework compliant with GDPR and the AI Act, REALM ensures a robust, ethical, and data-driven approach to decision-making in healthcare.

The REALM platform architecture is composed of five main components:

  1. Model Cards – Standardised documentation that captures essential information about medical device software, including performance metrics, ethical considerations, limitations, and regulatory approvals.
  2. Federated Cloud-Based Data Resources Catalog – A centralised, federated data repository containing real-world and synthetic datasets.
  3. Regulatory Toolbox – A comprehensive set of tools for testing, evaluating, and certifying medical device software, ensuring compliance with regulatory standards throughout its lifecycle.
  4. Living Lab Environment – A pilot space for real-world testing and evaluation of medical device software, focusing on user interaction and performance in healthcare settings.
  5. Post-Marketing Surveillance Module – An advanced monitoring system to ensure certified software maintains regulatory compliance and performs effectively post-deployment.

Capture

As we mark the second anniversary of REALM, here are the highlights and milestones that have defined its journey so far:

Building a Strong Technical Foundation
The REALM platform has laid a robust groundwork, combining advanced architecture with transparency-focused tools. Model Cards provide standardised documentation of software characteristics, ensuring transparency throughout the certification process. These are supported by a blockchain-backed system to enhance traceability.

To power its capabilities, REALM incorporates a federated cloud-based data catalogue that integrates real-world and synthetic healthcare data. By adopting the OMOP Common Data Model, the platform ensures compatibility with established health data standards, enabling seamless integration with networks like the DARWIN initiative by the European Medicines Agency.

Innovative Testing and Monitoring
REALM has launched five real-world demonstrators, validating the platform’s ability to assess medical device software in diverse healthcare settings. By integrating synthetic data and digital twins, the platform is equipped to handle complex testing scenarios, including rare diseases with limited real-world data.

Complementing these efforts is the development of a post-marketing surveillance module, which leverages RWD to monitor the long-term safety and efficacy of certified software, ensuring compliance even after deployment.

Ethical and Regulatory Considerations
REALM remains committed to ethical and regulatory excellence, collaborating with authorities such as the EMA and FDA to align with international standards. REALM has also developed recommendations for evaluating AI-based healthcare software, contributing to the upcoming ISO/IEC 63450 technical report on medical AI evaluation.

Fostering Collaboration and Stakeholder Engagement
Engagement with stakeholders has been the focus of REALM’s activities. The Multi-stakeholder Advisory Board ensures feedback from healthcare professionals, developers, and regulators, while the REALM Academy will offer different trainings for regulators on how to use the platform. REALM is also connected with other EU projects funded under the same call. Through collaborative efforts, the cluster aims to advance the field by exploring innovative approaches to leverage real-world and synthetic data for regulatory decision-making and health technology assessment. By improving the evidential value of these data and promoting their usability, the cluster contributes to improving regulatory practice and health technology assessment.

Looking Ahead
The progress achieved in these two years has laid the groundwork for a data-driven and ethical approach to medical device software certification. As the REALM consortium continues to refine the platform, expand its real-world demonstrators, and deepen its engagement with regulatory bodies and stakeholders, it is well-positioned to shape the future of healthcare regulation, enhancing trust, accountability, and patient safety.