What health, what data, what space? Three reflections on the announced European Health Data Space
1. EU strategy for data: making data sharing easy and meaningful
The EU considers that the ‘reservoirs’ that will power the sharing of data are data spaces. The idea has been in the works for some time; in 2018, the European Commission launched a communication entitled Towards a common European data space.
Central to its narrative is that an improvement in efficient use of data across the EU is needed; for this, “a seamless digital area with the scale that will enable the development of new products and services based on data” is proposed, encompassing the measures concerning re-use of public sector information, access to scientific information, and sharing private sector data. The notion that data is “an essential resource for economic growth, competitiveness, innovation, job creation and societal progress in general” is further advocated in the European strategy for data, rolled out in February 2020.
This is ‘the’ data strategy, aiming to bring about an EU single market for data where data can flow freely, the rules for its access and use are “fair, practical and clear”, and European values are respected. The strategy envisions, among others, the construction of a common market for data, which will be accompanied by the development of sectoral data spaces. This will be done by relying on data interoperability and portability to break the ‘silos’ and facilitate (citizen) control, and promoting more innovation and choice through businesses tapping into the potential of vast public sector data sources. The data spaces are intended to tackle the regulatory and technical challenges that stand in the way of realizing Europe’s potential in the digital(ised) economy and society.
1.1 Data space as a platform for better understanding of health and disease
Health is one of the sectors where data carries significant promise to boost our understanding of health and disease, especially considering recent technological advances. In this sense, improving and personalising health and healthcare through big data technologies is high on policy agenda; however, data is still very much fragmented across different systems that are often separate, incompatible, and scattered across Member States. To overcome this, the strategy will create the European Health Data Space (EHDS), alongside 8 other common data spaces tackling various areas of particular societal importance.
The creation of the EHDS is one of the priorities for the European Commission, as well as for the current Germany presidency of the Council of the EU. Its main objective is to facilitate the sharing of various kinds of health data across Europe, for both primary and secondary purposes, to support healthcare delivery, biomedical research, personalised medicine, prevention, diagnosis, and treatment of diseases, development of new treatments, products, and services, and overall innovation in the sector. The handling of data for doctors and researchers is now to be harmonised, which should ultimately contribute to better health.
This will be enabled by establishing a common code of conduct regarding the processing of health data, scaling-up cross-border exchange of data, including data from electronic health records, genomic, registries etc. through federated repositories, support for national electronic health records and data interoperability, cross-border exchange of electronic patient summaries, prescriptions, medical images, and lab results, and support for big data projects, all of course in compliance with the GDPR. The EHDS will come about by creating rules, ensuring data quality and interoperability, and supporting infrastructure. These steps also correspond to EU’s plans regarding the digital transformation of health and care and fit in the overall context of national governments digitalising their healthcare systems.
1.2 Data spaces governance under discussion: what matters are on the table?
The measures mentioned above correspond to the objective of facilitating free movement and harnessing data innovation. Efficiency of this kind is of course important and relevant in the health sector. But – is it the only thing the EHDS should be concerned with? The legislative framework to govern health and other EU data spaces is expected before the end of this year. In this light, I bring to attention some implications I hope will be contemplated in the governance of the EHDS. While the issues related to privacy and data protection usually feature most prominently in debates when it comes to sharing health data – which is only logical due to its sensitive nature -, here I touch upon other considerations that are just as relevant.
2. Which data is included and for what?
The mobility of health data in the EHDS encompasses, first, the support for efficient healthcare delivery to an individual patient when, for example, receiving a healthcare service in another EU Member State. This has direct and straightforward benefits both for the patient and the healthcare system when data “travels” across borders in care situations. Second, the EHDS sees using health data as a (raw) resource for innovation in the healthcare sector, including biomedical research, in an attempt to health records and other data sources from storage of information into “active pieces of the infrastructure underpinning medical research” Europe-wide. Reusing data for research is indeed to be welcomed, but the rules need to be clear. Further, data is not a homogenous input that captures reality in an impartial way; there are motives guiding its capture, selection, and processing.
This is especially relevant when data is “used in new contexts”. Health data assembled and shaped by differing national approaches does therefore not deliver automatically on the belief that “quantification produces truth”. In this sense, therefore, even if data from the health sector is abundant (and under-used), much effort and resources will be needed to curate and manage the data to be readily available for its uses, as well as to overcome the different national rules regarding access to data. The idea of the EHDS seems to make a step into that direction, so it will be interesting to see how the rules and principles are specified for and applied in the health sector (for example, what research purposes are actually included) and to follow the actual implementation and roll-out of the EHDS.
Moreover, as evident, the EHDS heavily focuses on data from the formal health systems and ‘conventional’ research, like electronic health records, registries, and repositories. What it seems to overlook – or leave aside for now – is another type of health data that could also contribute to healthcare: the emerging paradigm of digital phenotyping sees integrating the multidimensional data like the data from mobile and wearable technologies, apps or social media into clinical processes and even into health-related research as potentially useful and (positively) transformative. On the one hand, this could significantly contribute to what the EHDS has to offer, while, admittedly, also adding a level of complexity into an already complex field. Thus, with the growing wearable tech market and the quantified-self movement, such data should at least be considered in discussions about creating a data space for health. On the other hand, this points towards another development: such data often derives from technologies provided by private actors, including big technology corporations, which are entering the sphere of healthcare with full speed. Not only they procure (infra)structural elements, software and data management solutions, and specific products, like AI diagnostics or wearables, they also hold troves of health data themselves and develop new methods for data collecting and analysing or even provide health-related services. Often actors like Google, Apple or Amazon are portrayed as disrupting health innovation positively in delivering the common good, but, as pointed out, this has also raised rather significant ethical and other issues with several notable examples from the recent past. As mentioned, this element is rather brushed over in the strategy, which limits itself to ensuring the markets stay open and fair (more on that below).
Nonetheless, ignoring broader contextual developments like these risks that the increased role and power of large technology companies whose business model often revolves around data and AI will go unnoticed. This has proven contentious in the light of the current pandemic: it has been argued that through expertise and capabilities, these “public actors without public values” are taking on the role of health policy-makers. Case in point: Covid-19 contact tracing efforts by Apple and Google, for instance. The EHDS governance approach should thus not disregard this.
3. A market-based tool to sort out data sharing
But – what even is a ‘data space’? The term was proposed in 2005 as “a new abstraction for data management”. As such it contains all information that is relevant for a particular organization, “regardless of its format and location”. A data space relies on the notion of interoperability, which is “the ability to transfer and render useful data and information across systems, applications, or components”, meaning that different features (or services) can access these data either through technical solutions, like standards, or creation of data pools like repositories. Further, data portability refers to the capacity of an individual to ‘port’ their data from one service to another and is, among others, also expressed in the GDPR. Essentially, this technical solution, which is about linking and integrating the scattered and diverse data sources, is now used in efforts to build a common data market. The latter revolves around overcoming the (technical and legal or administrative) barriers to free movement of data across Member States and their health systems in order to promote individual control, enhancing choice and quality, more effective decision-making, dynamic competition etc. This is also reflected in the chosen legal basis for the EHDS: Article 114 TFEU. This article is intended for measures that aim to improve the functioning of the EU internal market by removing the obstacles to trade. This is only logical, as the EHDS is about facilitating structures and principles for data interoperability and quality to support its access, sharing, and reuse across the EU. While it is true that efficiency, choice, and competition resulting from a functioning market are indeed important features of ensuring quality and safety in healthcare, (access to) health is also about our other fundamental needs. I argue these values should play a more prominent role in the governance of the EHDS, as discussed below.
4. What health and for whom?
If the EHDS is ultimately intended to contribute to the use of health data for better healthcare, better research, and better policy-making, we should probably also ask ourselves how is this (data) contributing to overall better health. Data is an increasingly important element in production of knowledge, wealth, and power, including in the biomedical field. It is also importantly defining our relationship to and the meaning of health. Think only about self-tracking devices that enable individuals to me more involved in managing their health and lifestyle. Data and the patterns produced from it – like those showing our daily movement, sleep or reproductive cycles, or psychological wellbeing – are instrumental in how one relates to their health status; but this is an algorithmically-mediated relationship. What is the ‘health’ that data from the EHDS will actually define and produce, and for whom? Data-based health interventions and digital products often escape “scrutiny applied to other types of medical intervention or diagnosis” and without proper benchmarks regarding safety, efficacy, and effectiveness, it is hard to discern their quality from mere commercial opportunism. Further, it was also shown that AI systems developed from data can cement existing biases and reinforce inequities in terms of access to health. The EHDS should not turn a blind eye to this. While the strategy’s narrative seems to recognize the importance of using the data for social, aside from (only) economic good, the proposed measures are not really clear on how to go about that. The governance of EHDS should thus ensure that data spaces drive innovation that is of actual public value and operationalize how the EHDS can contribute to social protection, social justice, and underpinning values of universality, access to good quality care, equity, and solidarity – especially since the strategy itself claims that the measures should correspond to and promote European values.
5. Concluding remarks
Devising a data space for health is a noble and ambitious task. For now, though, it seems that the EHDS is more about data (market) and not so much about health. Governing EHDS by technical means will indeed ensure the market-like space for data is well-functioning, open, and a fair and level playing field. All this is important and necessary to deliver on the aims of the EHDS, facilitate innovation, and to improve healthcare. But data is not an end in itself; in order to make sure data-driven innovation in health will actually deliver benefits to society as envisaged, the EHDS governance should also consider ways to uphold norms and values central to (better) health. Let’s hope that the expected legislative proposal will deliver upon that!