The secondary use of health data for research purposes is set to rapidly escalate within the European Union with the introduction of the European Health Data Space (EHDS). The recently-agreed regulation seeks to regulate the secondary use of health data within the Union. Secondary use of health data refers to the use of health data for purposes outside its initial data collection. An example is the use of health data initially collected in the healthcare context to provide primary care to patients that is later used to do research. Secondary uses of health data can help researchers to undertake innovative research and ultimately improve healthcare outcomes for patients.
GDPR and the principle of ‘fairness’
Under the existing data protection regulation, the General Data Protection Regulation (GDPR), health data is classified as sensitive data, for which processing is only allowed under very limited circumstances. Indeed, the secondary use of health data is already regulated within the GDPR, albeit with many shortcomings such as unclarity over what ‘secondary use of data’ means and differing uses of the legal bases in the 27 Member States, among other things. This has created what some have called “(…) a divergent and fragmented [regulatory] landscape.”[1] Given the overlapping competences between the EHDS and the GDPR, the proposal has raised some questions regarding its compliance with the latter.[2] One of these questions is whether the secondary use of health data, as it is regulated in the EHDS, meets the principle of fairness. Fairness is one of the guiding principles in the GDPR, alongside lawfulness and transparency (Article 5 (1) (b)). Moreover, fairness is enshrined in Article 8 of the Charter of the Fundamental Rights of the European Union, where it is specified that personal data “must be processed fairly” [emphasis added].[3] In the context of scientific research more generally, fairness has been argued to be an important principle in research ethics, often closely linked to justice.
Fairness is enshrined in Article 8 of the Charter of the Fundamental Rights of the European Union, where it is specified that personal data “must be processed fairly”.
However, what a fair secondary use of health data means remains unclear. To better understand this, we performed a scoping review of the argument-based literature,[4] to determine how fairness for individuals is conceptualized in the specific context of secondary use research.
Four main factors were found to influence the understanding of what a fair secondary use of health data means for individuals:
- balance of interests
- power asymmetries
- commercial involvement and
- benefit sharing.
Balancing individual and public interests
Our findings show that striking a balance between public and individuals’ interests is very important in the context of secondary uses of health data. While it is clear that there are certain shared interests in health protection, as shown by the Covid-19 pandemic, not everyone agrees that a public interest in secondary uses of health data for research purposes can justify the removal of all control on data by individuals. What this control may mean in practice has been a topic of contention in both academic literature and in the trilogue negotiations, with a final agreement reached for an opt-out approach in the compromise text. This approach has also been supported by a number of NGOs and advocacy groups, such as the European Cancer Organisation.
Moreover, many authors argue that appeals to public interest need to be further substantiated. This requires providing an appropriate definition of what is considered to be in the public interest. The lack of a definition has previously been highlighted to be an issue in the context of the GDPR, where “ (…) public interest is mentioned 70 times, yet on none of these occasions is the concept fully explained.”[5] Given that in the upcoming EHDS public interest is introduced as both a legal and ethical basis for which electronic health data can be processed for secondary use, it seems important that future health data access bodies strive to have a common and shared definition of what types of secondary use research are in the public interest. Without explaining the reasons why such secondary use research is in the public interest, the EHDS risks creating an unfair scenario for individuals in which they are asked to expose their health data to higher privacy risks for unspecified public interest purposes.
Power asymmetries
Power asymmetries were also identified to pose a significant obstacle for the fair secondary use of health data for research. Therefore, in situations where individuals are not aware of their health data being used for secondary purposes or the data users are not clearly identified, this is unfair to individuals. This is because individuals who are not adequately informed about the use of their health data may be more vulnerable and susceptible to power imbalances. Some have claimed that this situation, also known as information asymmetry, may be aggravated by the EHDS,[6] given that the right to information under the GDPR is restricted under the proposal. Therefore, in order for secondary use research to be fair, these power imbalances need to be addressed. Engaging individuals by providing clear and transparent communication about health data use,[7] ensuring they are adequately informed about their rights[8] or involving individuals in the governance of secondary use research[9] are some of the suggestions put forward in the literature to tackle these power asymmetries. Moreover, the opt-out mechanism discussed above is also argued by some to empower individuals by giving them control over their health data.
Commercial involvement
Additionally, the involvement of commercial actors in secondary use research is also recognized to have an important impact on the discussion about fairness. Especially in cases where the commercialization of health data results in large financial profits and these are not returned to society, this is understood to be unfair to individuals. Despite these concerns, the involvement of private actors in the secondary use of health data is argued to be necessary to meet the goals and objectives of the EU’s Data Strategy, which strives to make “(…) the EU a leader in a data-driven society.” Given this reality, many in the literature argue that it is important to set conditions that make commercial involvement fairer to individuals. Suggestions include implementing accountability measures, such as imposing sanctions when data is misused,[10] or limiting the influence of large commercial actors on research agendas through the use of appropriate governance mechanisms.
Benefit sharing
The final factor that was revealed to be important to conceptualize fairness in secondary use research was benefit sharing. Benefit sharing requires that benefits derived from secondary use research, whether these be economic or not, are distributed equitably with individuals. The practice is argued to be an essential component for research ethics and is enshrined in the Declaration of Helsinki, stating “(…)if medical research is done within a vulnerable group, this group should stand to benefit from the knowledge, practices, or interventions resulting from the research.” Yet, in the case of secondary use research with health data, the importance of benefit sharing has generally been disregarded, partly given that many feel the risks to individuals are often insignificant.[11] However, our review shows that benefits sharing is considered essential to ensure secondary use research is fair to individuals. Some of the proposed benefit sharing measures include taxation (where profits are made from research),[12] securing affordable access to the results from research,[13] redistributing economic benefits with individuals or communities or ensuring capacity building.[14] Interestingly, in the context of the EHDS, demands for benefit sharing were put forward by a number of NGOs in a joint statement about the proposal published last year, in which they asked policy makers to include provisions on public return on data investment.
Looking to the future
It is encouraging to see that the compromise text of the EHDS has incorporated many of these concerns. Looking forward, however, a number of challenges remain in its implementation. For instance, how the opt-out will be rolled out in each country so that it is “accessible and easily understandable” (EHDS, Article 35f(1)) remains an important point of debate. Moreover, how the public interest of secondary health data uses will be determined by health data access bodies is still unclear and should be informed through public engagement activities and harmonized in the implementation process. Finally, it remains an important task for Member States to secure the necessary trust of society. Indeed, a lack of trust was manifested in the legislative process through a petition to include opt-in consent for the secondary use of health data in the regulation which collected over 100,000 signatures. A commitment to ensuring a fair secondary use of health data will be crucial to address these future challenges.
Author: Patricia Cervera de la Cruz, PhD Student at the Faculty of Law and Criminology (Ghent University) and member of the Metamedica-platform. Read the full article co-authored with Professor Mahsa Shabani.
References
[1] Slokenberga, Santa. “Setting the Foundations: Individual Rights, Public Interest, Scientific Research and Biobanking.” GDPR and Biobanking 43 (May 20, 2020): 11–30. https://doi.org/10.1007/978-3-030-49388-2_2.
[2] EDPB-EDPS Joint Opinion 03/2022 on the Proposal for a Regulation on the European Health Data Space. https://edpb.europa.eu/our-work-tools/our-documents/edpbedps-joint-opinion/edpb-edps-joint-opinion-032022-proposal_en
[3] European Union, Charter of Fundamental Rights of the European Union, 326 OJ C (2012), http://data.europa.eu/eli/treaty/char_2012/oj/eng (last visited Apr 28, 2023).
[4] Cervera de la Cruz, Patricia, and Mahsa Shabani. “Conceptualizing Fairness in the Secondary Use of Health Data for Research: A Scoping Review.” Accountability in Research 0, no. ja (2023). https://doi.org/10.1080/08989621.2023.2271394.
[5] Slokenberga, Santa. “Setting the Foundations: Individual Rights, Public Interest, Scientific Research and Biobanking.” GDPR and Biobanking 43 (May 20, 2020): 11–30. https://doi.org/10.1007/978-3-030-49388-2_2.
[6] Terzis, Petros. “Compromises and Asymmetries in the European Health Data Space.” European Journal of Health Law 1, no. aop (October 27, 2022): 1–19. https://doi.org/10.1163/15718093-bja10099.
[7] Linnet Taylor, What is data justice? The case for connecting digital rights and freedoms globally, 4 Big Data & Society 2053951717736335 (2017);
[8] Gemma A Williams & Nick Fahy, BUILDING AND MAINTAINING PUBLIC TRUST TO SUPPORT THE SECONDARY USE OF PERSONAL HEALTH DATA (2019).
[9] Alessandro Blasimme, Effy Vayena & Ernst Hafen, Democratizing Health Research Through Data Cooperatives, 31 Philos. Technol. 473 (2018).
[10] Aisling McMahon, Alena Buyx & Barbara Prainsack, Big Data Governance Needs More Collective Responsibility: The Role of Harm Mitigation in the Governance of Data Use in Medicine and Beyond, 28 Medical Law Review 155 (2020).
[11] Porsdam Mann, Sebastian, Julian Savulescu, and Barbara J. Sahakian. “Facilitating the Ethical Use of Health Data for the Benefit of Society: Electronic Health Records, Consent and the Duty of Easy Rescue.” Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 374, no. 2083 (December 28, 2016): 20160130. https://doi.org/10.1098/rsta.2016.0130.
[12] Barbara Prainsack et al., Data solidarity: a blueprint for governing health futures, 4 The Lancet Digital Health e773 (2022)
[13] Luca Marelli, Giuseppe Testa & Ine Van Hoyweghen, Big Tech platforms in health research: Re-purposing big data governance in light of the General Data Protection Regulation’s research exemption, 8 Big Data & Society 20539517211018784 (2021).
[14] Viroj Tangcharoensathien, Jirawan Boonperm & Pongpisut Jongudomsuk, Sharing health data: developing country perspectives, 88 Bull World Health Organ 468 (2010).