Optimising cohort data in Europe

Moreover, in emergency situations, people are more willing to give away their personal and confidential data (place of living, psychological needs) or their rights to privacy and data monitoring/access (Merchant and Lurie, 2020). Online contexts exacerbate this problem because they offer structured choices that ''force'' participants to consent (Chancellor et al., 2019). The bargaining power of participants in the context of mobile data collection is thus limited because they have little to say on how their data will be shared and used (Hand, 2018). The same phenomenon applies to vulnerable populations where the need for problems being solved far outweighs any concerns about privacy (Copes et al., 2018). Mobile data collection has no real safeguards against the misuse of personal data in such circumstances, which means that private, intimate data collected during emergency times or from vulnerable populations can potentially be used later in a discriminatory manner (e.g. health benefits) (Ali et al., 2019). Social media data data collection technologies are increasingly tied to personalise medicine. It is suggested for instance, that patients' social media data and healthcare records could be combined into a personalised profile that could be accessed by healthcare providers later on (Hunter et al, 2018). However, such practices may harm participants' autonomy and privacy rights. Collecting social media data in the context of clinical environments thus requires a familiarity with the legal and ethical frameworks in place on the one hand, and a recognition that patients are increasingly involved in their own care on the other hand (Vayena et al., 2018). In other words, patients' aspirations and interests for a more active participation can be successfully implemented through research but only within existing legal and ethical frameworks (Taylor and Pagliari, 2018). The ethical issues of solidarity and commitment to the rights of vulnerable communities take a crucial meaning when recruitment through social media (e.g. online survey) is concerned. There are two important challenges in this context namely, the challenges of self-selection and representativeness (Arigo et al., 2018). In the former, some types of social media users are recruited at the expense of others (Khazaal et al, 2007). In practice, this means for instance, that only participants who manifest an interest in the study because of their occupation or/and professional profiles take part. In the latter, there is a risk that participants with higher socio-economic status will be recruited at the expense of other social categories (Copes et al., 2018; Arigo et al., 2018). Namely, sustained online access (including for online surveys) requires literacy, a regular internet access and a willingness to identify oneself online. All these characteristics are rarely present in underserved and minority populations, which suggests that they can be under-represented in these kinds of studies (Gostin et al., 2018). In this sense, reaching vulnerable communities through social media should be bound with a commitment to their safety, privacy and security. Passive data collection raises important challenges to autonomy because the ambiguity of the secondary data use. That is, the secondary use of passive data is not straightforward