These days I am completing a Microcredential in European Health Data Space (EHDS) and I recorded a summary video of what I have learned so far. Here it is.
Summary video: what EHDS is and what it means for clinical research
Beyond the course content, the EHDS has given me a lot to think about and I wanted to share some of those reflections.
It is undoubtedly good news that a regulatory framework exists to guarantee data interoperability across EU member states for primary use of data. This lack of interoperability directly affects the quality of patient care, and having a common strategy is a necessary and welcome step.
That said, I am concerned that just OMOP seems to be consolidating as the standard in secondary use. OMOP is primarily designed for observational studies, and while the EHDS discourse mentions genetic data and precision medicine, there is still no clear strategy in that direction: in terms of which standard will be used for secondary analysis in precision medicine, nor how patients' genomic information will be coherently integrated.
Leaving genetic information properly modelled out of the strategy for secondary use would be a mistake. And relying solely on raw data transfer, without true semantic interoperability, would be a step backwards from what the EHDS promises: competitiveness with the US and China, and cross-border semantic interoperability without knowledge ambiguities between countries. The chosen standard should not be FHIR for secondary use (just as it has not been the choice for clinical data because of technical reasons that we are not going to discuss here) but we should not settle for less than what science needs either.
Finally, there is the question of consent. Ensuring that all patients in member states have the right to rapid and equitable access to the benefits of new discoveries and treatments is something the EHDS does not guarantee explicitly. In industry, we are taught that data is a very high-value asset. With the EHDS, consent will no longer be opt-in by default.
I understand and support this step, as long as immediate value return to the patient who has shared their data is guaranteed. Altruism and long-termism, when talking about patients suffering diseases, should always come with explicit consent unless the patient is likely to obtain some benefit in return. Otherwise, we risk building a system where value flows in one direction and trust is broken in the other.
I will keep sharing more reflections as I progress through the Microcredential. If you have thoughts on any of these points, I would love to hear them.