Big data could drive massive improvements in European health systems, but only if we can agree on a set of rules for the use, reuse and management.
By Pierre Meulien, IMI Director
There is no doubt that the ‘big data’ challenge is nowhere greater than in the healthcare sector. Big data refers to the huge volumes of data from multiple and diverse sources, such as health records, medical imaging, genome sequencing, research, wearables, medical devices and so on. The data are being generated at ever increasing speed, and need to be managed in a coordinated fashion. They are potentially extremely valuable, and could drive massive improvements in European health systems, but this will only become reality if we can agree on a set of rules for the use, reuse and management of the data.
Harmonising complex datasets, whatever their source, making it easier for them to be ‘found’ by machines, making them interoperable (i.e. capable of being integrated with other datasets or applications), and encouraging their reuse, not only reduces duplication in research efforts; it allows us to raise the bar on the quality, as well as the equality, of healthcare across Europe. The Innovative Medicines Initiative (IMI) has been engaged in this data revolution over the past decade.
Defragmenting patient-level data
To make data ‘FAIR’ means to make it findable, accessible, interoperable and reusable. The guidelines were drawn up by different actors from academia and the private sector in 2014 as a way to overcome obstacles in data discovery and reuse. An early IMI project, Open PHACTS, was involved in the drafting of the original influential manifesto. IMI has adopted the principles across our project portfolio, starting with the project EMIF, which resulted in a platform that is helping to defragment patient-level data. A concrete example of how the project has made data ‘findable’ is a freely accessible data catalogue, listing data on Alzheimer’s Disease, cardiovascular diseases and vaccine safety. The project has created tools to efficiently access and use that data by putting it in an ‘interoperable’ common format.
Improved treatment protocols
Accessing standardised data at scale can also improve the current treatment protocols. For example, HARMONY, one of our big data projects, looks at standards of care in treating rare haematological tumours like leukaemia. Due to the small numbers of patients in rare diseases, it can be difficult to understand which treatment is best for an individual patient. By gathering and ‘FAIRifying’ data from over 45,000 patients across Europe, HARMONY is quantifying the clinical outcomes of particular treatments, in order to decide on the best standards so that they may be implemented right across the continent.
Video: IMI is promotes the FAIRification of data from our projects - it not only reduces duplication, but brings up standards of care across the EU.
Equality of care, wherever you are
This means that citizens can expect the same standard of care no matter where they are in Europe. It is of immense value to those who are in charge of public health policies and budgets, whether it’s from a clinical decision making point of view, from an economic choice point of view like helping the technology assessors or public health interventions and policy. The same research goal is being advanced in other therapeutic areas such as cardiovascular disease (BigData@Heart project), Alzheimer’s disease (ROADMAP project) and prostate cancer (PIONEER project).
The ‘European Health Data and Evidence Network (EHDEN) is a project that is putting these principles into practice at scale. EHDEN aims to make the data from 100 million European citizens available for research, while fully respecting patient privacy. To address the current COVID-19 outbreak and have prioritised the data of COVID-19 patients and have already published the biggest ever study on the safety of hydroxychloroquine, which is already being assessed in some COVID-19 patients.
Towards a new culture of data management
Adopting FAIR principles is also good for business. In 2018, the European Commission published a report warning that not having FAIR data costs the European economy at least €10.2bn every year. Of course, it can difficult to know how to apply these principles of data management, which is why we launched FAIRPlus in 2019. FAIRPlus looks across the board within the IMI portfolio to try to ensure the guidelines are adopted.
The biopharmaceutical sector is a big producer and consumer of data, they recognise that opening up their data – such as in a pre-competitive space like that offered by IMI - will help break down barriers that are impeding the development of new treatments and drugs. Yet many organisations are still reluctant to apply the FAIR principles and share their datasets due to existing cultures and behaviours. FAIRPlus is organising training courses for data scientists in academia, SMEs and pharmaceutical companies. Ultimately, the project hopes to change the culture of data management in the life sciences sector.
Read more
From a few clicks to a few months - FAIRified datasets can fast-track drug development
More and better patients data, whatever the disease