The overarching goal of the EHDEN project is to foster a thriving European community that better shares health data research, and in particular, to enable federated analyses of real- world healthcare data to ultimately improve patient treatment and outcomes. By definition, this is not something that the project partners can do alone: rather, we see our project as a catalyst to come together and build this with hundreds of other projects and organizations. And indeed, through the EHDEN Academy, our Harmonization Fund grants and the SME training programme we have already worked with thousands of people throughout the world to broaden this community, working with OHDSI, sister projects such as PIONEER, regulatory agencies, etc.
However, there is another side of the coin, a particular driving force for EHDEN sine qua non: the use of open standards and open source software. It’s true that by itself, a standard does not mean much, as illustrated by this well known XKCD comic about standards. But combine it with a thriving global community of people and organisations that develop, maintain and implement the standard, and you then have what EHDEN project lead Nigel Hughes calls a ‘socio-technical construct’.
Another way to look at it, through the lens of the ReEIF model, is that in order to achieve interoperability between healthcare organisations, research networks, disease registries, patient applications, etc., alignment is needed on all levels: legal, policy, process, information and data, application and infrastructure levels. It does not mean that all organisations need to adopt the exact same policies, processes and standards, but if they are misaligned, then the collaboration is likely hard to sustain. And sustainability is another very important goal for the EHDEN project: we are all in this for the long run.
So let’s have a look at where EHDEN stands today in terms of data models, which is an important aspect of enabling interoperability of information and data. Recently, we published EHDEN D4.5 Roadmap for interoperability solutions, which covers this topic in detail. Of course, the OMOP Common Data Model (see below) plays an important role in the EHDEN project, because this is the basis of how we harmonise healthcare data across Europe with the EHDEN Data Partners. However, OMOP is not the only standard that is relevant here. A growing number of healthcare organisations are implementing HL7 FHIR internally for interoperability of their systems, and momentum to use FHIR for research purposes is growing. So conversion from FHIR profiles used in practice to OMOP for standardised analytics is a hot topic for EHDEN (and for OHDSI) and several of our Data Partners are building this. And although not in scope for EHDEN, other projects such as CINECA (and GA4GH) are building out similar models for genomics data, imaging data, wearables data, etc.
Finally, there is the important topic of ‘metadata’, defined as ‘data about data’ or ‘data descriptors’ (although this definition can be a bit of a rabbit hole sometimes). In EHDEN, two of the most important forms of metadata are information about all the databases available in the network, and the medical studies that are designed, run and published in the network. In D4.5, we propose an adaptation of the widely used schema.org model to advertise this metadata, in line with other projects such as Bioschemas. However, this is clearly an important area where alignment with other national and international initiatives such as HDR UK, All of Us, MII, SPHN, Health-RI (to name just a few of many health data networks), as well as with regulatory agencies such as EMA and FDA would be extremely beneficial. EHDEN is actively engaged in conversations with many of these stakeholders, and we welcome any input from you on this important topic for open science in the long run!