Translational research is a promising approach to speed up discovery of new therapies and diagnostic methods. In order to realise such objective, tight collaboration of biomedical researchers and clinical practitioners is required [
1,
2]. Their work is data intensive [
3,
4] and must rely on information technology to enable efficient data exchange and analysis [
5,
6]. Compared to more traditional drug research, access to a larger variety of trials from diverse sources can improve the characterisation of benefits and unwanted effects of drugs and therapies at lower costs and better efficiency [
7]. Drug approval processes and drug safety/effectiveness surveillance are improved by faster access to data about active ingredients similar to the ones being under consideration. An example of that is the effectiveness of using existing evidence, or even the prior obligation to make trial outcomes publicly available, to prevent selective reporting [
8,
9], that is, the presentation of evidence that is favourable for the interest of the reporter (such as having a drug approved), and the exclusion of unfavourable evidence. Another potential advantages of these approaches is making clinical experimentation more efficient and avoiding the exposure of trial’s potential participant to known risks, as well as, for instance in the case that evidence shows adverse effects to particular health conditions, avoiding unnecessary risks. Ability to perform data analyses other than those for which clinical trials were originally conducted is another opportunity that clinical data sharing offers [
10], which is relevant in the translational research field, enabling approaches like comparative genomics [
11,
12]. Overall, this has social benefits such as faster improvement of healthcare and its safety, and increasing the confidence of the general public in the scientific community, public services and industry [
10]. On the other hand, dealing with biomedical information, and with human patient data in particular, poses complex challenges with respect to ethical, legal and social implications (ELSI [
13,
14]), which need to be addressed when software products are developed and IT infrastructures deployed [
15,
16]. An obvious example is the wish and right of patients to keep their health information private, which can be motivated by various reasons, including the kind of relationship that an individual wants to maintain with his relatives and social relations [
17,
18], the social stigma associated to certain diseases [
19,
20], and access to private healthcare [
21]. Another reason to resist data sharing lies in the commercial or academic interests of researchers, including the willingness to be the first to submit unpublished research, and the wish to produce evidence useful to file patent applications [
22,
23]. These issues pose potential conflicts with the research needs. For instance, anonymization and reidentification-prevention techniques, which are used to grant data access while ensuring patient privacy, imply that data essential for a research goal might be concealed from the researchers [
24,
25].
Life science shares technological challenges with other areas of science [
26], and generic technological solutions can be employed, either of commercial or open source type [
6,
27‐
29]. However, addressing ELSI in the translational research arena is particularly difficult, due to the above mentioned reasons, which can be summarised as heterogeneity of information systems, different types of professional roles involved, the conflicting needs to share information and, at the same time, ensure this is done in a way that respects patients and associated legislation [
30‐
32]. The domain of biobanking and biosample research is characterised by special restrictive sample and data usage conditions, since highest ethical standards to ensure the support and participation of human research participants are required. In addition to confidentiality, consent about the data usage, intellectual property and data/sample ownership must be considered. Sophisticated mechanisms to provide restricted access to sensitive data is a way to address this problem. The risk of improper use of the data can be mitigated through legally binding agreements, subscribed by trial participants and researchers, which constrain the purpose for which data access is granted. Access is mediated by some form of a data access agreement between a data consumer and a data provider. These access agreements have to take into account legal and ethical requirements, professional guidance, and good practices. Agreements are in general executed by data stewards or data access committees, but recently they are implemented in electronic form employing software for identity and access management. This approach is not without difficulties, such as the impossibility to foresee useful research goals at the time of data and consensus collection [
9,
10]. However, it can be seen as a compromise between the different needs that it addresses.
In this paper we report on a pilot implementation (from now on, ‘the pilot’) that aims at integrating research resources and clinical resources, including data bound to a varying range of access policies, from fully open to data requiring access approval. Implemented in the context of the BioMedBridges project [
33‐
35], the pilot shows how identity and permissions management can be simplified by means of a modular approach, utilizing well known software components.
The BioMedBridges project
The European Strategy Forum on Research Infrastructures (ESFRI) initiative has been promoting an agenda to build Research Infrastructures (RIs) in Europe since 2002 [
36]. Its current agenda comprises 21 projects in all scientific fields. This includes RIs for the life science area, several of which teamed up in the FP7 BioMedBridges project. The main aim of this project was to facilitate the translation of ideas into medical applications, by promoting data interoperability in a variety of disciplines, across different scales. The project concentrated on five use cases, including cross-species data integration, personalised medicine, imaging, and structural biology. This work was supported by technological, cross-domain activities, such as terminology and data standards harmonisation [
37], and secure access to data. The latter was investigated both from the point of view of ELSI, as well as what concerns the realisation of concrete IT solutions. All reports of the project are available [
38]. The pilot presented here is documented in detail in the report D5.4 [
39], which was preceded by the analysis and design done for D5.3 [
40] and by the preparatory investigations on ELSI topics in D5.1 [
40,
41] and D5.2 [
42].