Optimising cohort data in Europe

In a big data context, the value of biobanks is not dependent on the participants' perspectives on data use but rather on their statistical power (Richterich, 2018). The priority for biobanks is to be integrated into large networks that allow combining different data sources. Thus, it is questionable that patients' participation could bring anything more than surface sustainability to data-sharing processes. Namely, researchers have established strategies for ensuring privacy and security during data transfer that structurally and operationally do not afford integration of participants' autonomy (McRae et al., 2020). Participants' integration and the respect for their privacy and security rights becomes thus a purely technical matter. This means that participants' real contribution and influence do not go beyond consent arrangements (Nurmi et al., 2019). (b) Motivation for data sharing: the place of researchers Another important issue is why researchers share the data in the first place. Data sharing is framed as a contribution to the public good or as an expression of the right for free inquiry (Cech, 2019). However, issues of free inquiry and the public good are not straightforward because various types of stakeholders differ in their claims for data sharing and use. For instance, institutions and biobanks have fiduciary rights: they have full data stewardship including the control, the distribution, use and disposition of data (Juengst and Meslin, 2019). In this context, researchers are disadvantaged by (i) their lack of resources in preparing data for biobanks (Manrique de Lara and Pelaez-Ballestas, 2020) and (ii) the current lack of stable credit attribution mechanisms (Boeckhout et al., 2018). As a result, there is no clear credit attribution for data collectionwhile there are established credit systems for data storage and data analysis in biobanks (Coppola et al., 2019). Extended collaboration networks in the precision medicine domains is translated into increasingly longer authorship lists. This creates challenges in determining intellectual contribution especially for those researchers who were involved in data collection (rather than data analysis) (Hulsen et al., 2019). These issues are exacerbated by the legal fragmentation within the EU regarding privacy rights in general and GDPR implementation in particular (Townend, 2018). For instance, there is considerable uncertainty around data transfer because EU countries can implement additional own rules for data processing and protection, thus limiting GDPR's support for “free” data flow (Molnár-Gábor and Korbel, 2020). The rationale for sharing data is thus rarely determined by researchers but rather by grant terms, funding, peer-review requirements, country regulations and ethical committees’ approval (Kiehntopf, 2019). Uncertainties in data sharing across EU countries can be partly solved through sector specific rules (codes of conduct) (Molnár-Gábor and Korbel, 2020). The development of codes of conduct is supported by the GDPR and relies on the assumption that specific data processing fields have different needs. However, the positive impact of codes of conduct on the harmonisation in data sharing materialises only in the long term, which suggests that these cannot be implemented in isolation (Pastorino et al., 2019). Hence,