We aim to transform the mostly population based management of Multiple Sclerosis (MS) of today into an individualized, personalized and precision level management. We believe the key to achieve this next level of MS management is “F.A.I.R.” data. F.A.I.R. stand for Findable, Accessible, Interoperable and Re-usable. Currently, data is used rather superficially whereas we assume many new insight are to be discovered using the data that is already there. However, two main hurdles obstruct us from reaching these insights. First, most data is not findable, not accessible, not interoperable and not re-usable. Secondly, we lack proper analytical tools for optimal data mining. With this project, we aim to overcome these two obstructions.
Imagine any type of data being ‘Findable, Accessible, Interoperable and Reusable’ by both humans and machines. Important to note here is that “findable” does not mean “for everyone to find”, “Accessible” does not mean “open access”, “interoperable” does not mean “for everyone to operate on” and “re-usable” for everyone to use”. However, it creates the possibility to find, access, interoperate, and re-use data when necessary. In other words, it gives data the opportunity to have maximal impact. The possibilities to discover new insights multiplies manifold. We may finally be able to make far greater strides forward than ever before in the challenging MS management.
But before we get there, many hurdles have to be overcome. There is an urgent need for data collection procedures and tools that create F.A.I.R. data. Data isn’t FAIR because you open it up to others. We should all become more disciplined in collecting and storing data the correct way, i.e. the F.A.I.R way. This project investigated these hurdles, opportunities and challenges in the concept of MS.
We aim to develop 1° data collection procedures and tools to create data that is F.A.I.R. (=findable, accessible, interoperable and re-usable), 2° IT solutions to allow (temporarily) pooling and linking of F.A.I.R dataset, 3° statistical methods to define minimal requirements for datasets and 4° new analytical methods for optimal mining of connected and pooled F.A.I.R datasets. Two pilot studies are performed.
These work packages are implemented on 3 different levels: 1° A local level: a proof-of-concept study in which retrospective data from 6 partners involved in care, rehabilitation and research are connected into a multidisciplinary data infrastructure; 2° A national level: in collaboration with the healthdata.be platform, a Belgian national multidisciplinary MS repository is developed using FAIR data collection procedures and IT solutions to allow secure data pooling of data prospectively collected by different stakeholders; 3° An international level: we are actively involved in the MS Data Alliance, a project on better exploitation of real world evidence data driven by the European Multiple Sclerosis Platform (EMSP).
We have the unique opportunity to collaborate with the healtdata.be platform. Healthdata is an initiative of the Belgian Scientific Institute of public health WIV-ISP to simplify the registration and storage of health care data in Belgium. Healthdata and the MS DataConnect consortium will collaborate to set up a multidisciplinary MS register combining information collected by care givers, patients and researchers.
A collaboration with healthdata enables us to create F.A.I.R. data collection tools and procedures for MS relevant data. The key principle of healthdata is the “only once”-principle referring to 1° data capture from primary (operational) sources of health care actors and 2° re-use of previously collected data. The generic HD architecture is approved by the Sectoral Committee of Health (privacy commission) and the e-health platform, guaranteeing privacy and covering ethical concerns.
The figure below provides a schematic overview of the project plan.