5 August 2015
From 30 June to 2 July, RD-Connect Registries work package leader Domenica Taruscio and her team at ISS hosted the second “bring your own data” (BYOD) meeting in Rome. This meeting welcomed registry owners, biobank owners and data linkage experts with the main aim of the meeting to move closer towards the goal of linking data from these varied sources.
The meeting started with an update from Sabina Gainotti and Paola Torreri (ISS) on ID-Cards, the online searchable catalogue of patient registries. In addition to the phenotypic and genotypic data collected, this system also includes detailed data on the standards, structure and governance of each registry. All registry owners are encouraged to provide as much information as possible to help make this system a valuable resource for the community.
Currently the information on the number of cases recorded in registries is uploaded manually by registry owners inside the “disease matrix” and one main aim of the workshop was defining a standardised procedure for the automatic update of the disease matrix through application programming interfaces (APIs). Robert Reihs from the University of Graz is defining the procedure for registries participating in the ID-Card system and there was a lot of discussion among IT experts around the best API solutions.
Marc Hanauer (Orphanet) presented the work Orphanet has been doing to match rare disease terms with Human Phenotype Ontology (HPO). Great progress has been made, with 2050 diseases annotated to 4500 HPO terms; the total process is expected to be completed by the end of the year. This annotation/ontological theme was carried throughout the two days with an emphasis on the need for registries to consider how they are collecting and coding data in order to make it easier for them to attain interoperability in the future.
Agata Robertson (Newcastle) presented an update on the RD-Connect plans for a unique identifier for individuals whose data is held within participating registries, biobanks and the RD-Connect omics platform. To be known as the RD-ID, this aims to ensure it is possible to establish whether particular data items belong to the same individual without identifying the individual or requiring any personal information to be disclosed. This enables data linkage and prevents duplication of data, while protecting patient confidentiality. The system proposed will be unique to RD-Connect but use data items that are interoperable with the HD-ID and ongoing NIH projects. This work is in progress, and the registries in the Core Implementation Group (CIG) are encouraged to see if they collect the appropriate items and if not to consider the resources needed to collect these both prospectively and retrospectively.
Over the remainder of the meeting Marco Roos (Leiden) led a number of breakout sessions focusing on the various hurdles facing the linking of biobanks and registries. These covered aspects that are important from both the user perspective, e.g. core data elements and a glossary of terms and the technical point of view e.g. data access, data flow and APIs.
One group focused on the work Manuel Posada and Estrella Lopez (ISCII) have been doing surrounding common data elements. A clear list of 35 elements has been produced and discussions surrounded what incentive there should be for registries to collect these elements and what, if anything should be mandatory. Again it will be important for the CIG registries to assess the data currently collected and if this could be adapted in the future.
The data linkage experts used test data exports provided by the European Huntington Disease Network, Global FKRP Registry and UK FSHD Registry in an excel format to develop a proof of concept. This allowed one group, led by Mark Wilkinson (Madrid) to establish a first draft using a FAIR assessor on top of these exports which uses an API linked data platform. The data linkage experts will develop this method further and test its usability on the registries and biobanks involved in the CIG. It became clear that the expertise for the technical solutions are available within RD-Connect, however the challenges remain in the access to data on a governance level e.g. are the right consents in place and the need to translate these achievements into lay language for users to understand.
In summary, the outcome of the meeting was very positive and concluded that it will be possible to link all of the data on some level. However, it is now important for registries and biobanks to consider how they need to adapt to these changing technologies and for RD-Connect to consider how to provide incentives for them to do so.
This work is ongoing and a follow up meeting will take place in Rome, 24-25 September, the RD-Connect Workshop on Data linkage and ontologies. This follows the 3rd International Summer School on Rare Disease and Orphan Drug Registries.
A full meeting report will be available soon.
~ Libby Wood, Curator of the UK Myotonic Dystrophy and FSHD Registry Coordinator