Optimising cohort data in Europe

because research proposal and frameworks become increasingly technical (Resnik, 2019). The focus is not much on secondary use per se but rather on the platforms, algorithms and diagnostic tools which make this secondary use possible both in legal and in practical terms (Corwin et al., 2019). In this context, institutional review boards experience increasing difficulties not only in determining the scientific worth of the proposals but also in evaluating the potential risks and benefits of participation (Janeček, 2018; Di Matteo et al.,2018). The increasing technicality within the passive data collection and analysis domains is also confusing for the participants themselves and restricts their autonomy in the study. In particular, participants have difficulties in determining the significance of their contribution or the potential benefits of their participation (Segura Anaya 2018; Nebeker et al., 2019). There are strong indications that researchers still don't have a clear strategy for fully integrating participants within the research process. Geospatial data collection technologies challenge autonomy principles because participants are relatively powerless with regards to the protection of their own identity and potential confidentiality breaches. There is also an increasing difficulty in educating participants of the extent to which geospatial devices and applications can reveal highly personal data (i.e. participants tend to accept any conditions for using commercial apps without giving much thought on what they entail) (Cetl et al., 2019). 3.2. Confidentialtity and privacy Confidentiality issues around mobile data collection revolve around the aggregation effect. Namely, data intensive research through mobile data usually combines various data feeds together (Ali et al., 2019). This combination is conductive to potential incidental disclosures about participants' information and personal data (Crawford and Schultz, 2014). Moreover, the tools and technologies for combining data are continually evolving which means that potential incidental revelations are impossible to prevent and predict. For instance, while revealing a specific piece of information may not disclose sensitive data by itself further potential combinations are likely to reveal highly personal information (Cornet and Holden, 2018). Data collection through social media may harm confidentiality rights through data sharing. That is, the potential for revealing information through data combination requires serious considerations of the contexts where this information is shared. For instance, data collection through social media of patients-doctors interactions involves numerous legal and ethical challenges. Patients' records are protected by law (e.g. the Data Protection Directive 95/46/EC in Europe) but communication and monitoring in healthcare are increasingly outsourced to social media (e.g. some medical appointments are made online) (Benetoli et al., 2018). Moreover, in the context of Internet-based communication, privacy settings are not absolute because the internet in general, and online platforms in particular, record content permanently (Chancellor et al., 2019). The use of social media in patient-doctor communications can also reveal patients' information that would have