By Mario Romao, Director, Health and Data Policy
At the 2nd European Alliance for Personalised Medicine (EAPM) Congress, held in Milan 26-28 November I had the pleasure to speak at a session on health data sharing. Whilst doctors are the ones saving lives, data – and the information thereof – are the building blocks of medical knowledge. In the era of Artificial Intelligence and precision medicine, access to quality data has been brought to the forefront of the healthcare debate.
Results from “HIMM’s Annual European eHealth Survey” reveal that 73% of patient data is reported as digitized and that trends for the near future will be: patient (owned) health records, health information exchange with external providers, patient self-monitoring initiatives, Electronic Medical Records implementations, artificial intelligence and personalized medicine.
These trends align with the priorities of this year’s European Commission Communication “on enabling the digital transformation of health and care in the Digital Single Market” which, among other measures, aims to promote cross-border exchange of patient records by issuing a Recommendation on technical specifications about how citizens’ electronic health records can be exchanged across the EU and the exchange of genomic information of at least 1 million individuals in support of precision medicine by creating common technical specifications to support secure cross-border access to genomic and other health data for research purposes.
The digitization of patient records, important as it is, doesn’t tell the whole story. Which kind of data is effectively in store, in which format and how reusable it is, are all determinants of an effective data -based ecosystem allowing for greater strides in understanding and preventing diseases, improving personalised medical research, diagnosis and treatment.
According to a survey from the European Commission, 80% of Europeans would agree to share their health data (if privacy and security are ensured). Whilst patients seem willing to share their health data, the extent to which that does not happen for clinical and research purposes is well documented. For example, the European Commission, in a Staff Working Document refers that “it is widely shared among the stakeholders that access to varied data-sets located across different Member States remains difficult or inexistent, the data is subject to different taxonomies and standards and therefore scientific research invariably builds on relatively limited population cohorts”. And a recent analysis of data sharing guidelines issued by multiple organisations “observed substantial fragmentation in the policy landscape across specific organizations and data types” and that “data are not being shared to a degree that can trigger the expected data-driven revolution in precision medicine”.
The reasons for suboptimal use of health data are multiple. A recent report focusing on oncology data sources in Europe, highlights challenges related to limited health data digitization, quality, format and accessibility; lack of sustained political focus on health data; heterogeneous and time consuming patient consent frameworks; limited technical interoperability and concerns about data sharing by healthcare providers themselves.
To accelerate data driven innovation in healthcare that meet the expectation of patients, healthcare providers, researchers and governments, Intel recommends to:
-Make health data digital and FAIR
-Consider codes of conduct to drive trust
-Strive for regulatory alignment
The EU should aim to achieve 100% of health data digitised within the next 5 years. This data must be Findable, Accessible, Interoperable and Reusable. Following FAIR data principles as a basis for the governance of health data, means that health data is stored according to widely accepted standards to facilitate its search; that technical and organizational measures are in place to securely provide access to the data; that the quality of the data follows widely agreed metrics and standard formats and that attribution is assured so that incentives and rewards for data sharing are in place.
In many cases, anonymizing personal health data will detract from its utility. In Europe, processing of data concerning health is regulated by the General Data Protection Regulation (GDPR). As the European ecosystem for health data develops, so will data collection, collaboration amongst stakeholders and opportunities for data processing. In this context, concerns with privacy and data protection will equally increase. Codes of conduct are a means to ensure the proper application of the GDPR, reinforcing trust by demonstrating compliance to the public and regulators. Within the context of healthcare research, for example, codes of conduct would contribute to network effects because their increased adoption by stakeholders would lower the barriers to enter in large scale collaborative research initiatives.
While the GDPR is an EU-wide data protection law, it does not completely harmonize data protection rules across the EU. In particular, individual EU Member States are allowed to adopt national rules that “supplement” the GDPR, for example, in relation to the processing of health or genetic data. It is thus very important that Member States and the European Commission minimize, or even better, eliminate, regulatory divergences regarding data concerning health.
In the era of precision medicine and artificial intelligence, we need to get the policies and practices right to extract the maximum from data, to improve clinical care, to drive new therapies and treatments, to make healthcare systems more efficient. Because, in the end, data saves lives.