Optimising cohort data in Europe

3. Pillar II: Guidance - Guidelines of harmonised legal, ethical and organisational conditions for access Specialised tangible resources dominate Pillar II because guidance inmethodological, legal and ethical terms requires explicitly quantifiable measures and fields of application. Pillar II is aimed to provide guidance to researchers and institutions with a range of different backgrounds, affiliations and interests and as such, needs to be as explicit in its advice as possible. The dynamic between specialised and versatile resources differs in significant ways from Pillar I. In Pillar I (i.e. standards), versatile resources are nearly as important as specialised ones. Versatile resources and specialised resources can be converted into each other because they have the same weight throughout the standardisation process. In Pillar II however, versatile resources decrease in importance. The components of Pillar II include: 1) federated infrastructures & analysis components, 2) broad consent platform, 3) trust temporal checkpoints, 4) meta-content model, 5) path-oriented governance for consent, 6) intersection domains for confidentiality (governance structure and federated analysis), 7) interim metrics of impact for research benefits, 8) sustainable accountability mechanisms, and 9) quantitative measures of data use. In order to achieve the predominance of versatile resources, we will use federated infrastructures and trust checkpoint components (i.e. components for monitoring the researcher-participant relationship, including trust) for illustrative purposes. The capacity for federated infrastructure and federated analysis does not depend directly on versatile resources but rather on specialised, tangible ones. However, the degree of the specialisation and specification of resources is important. For instance, costs and funds resources for federated structures are unstable because they present potential resource position barriers. That is, it is not always possible to determine who will assume the costs, pay for the software or recruit managers for using the software on the national nodes. Even if the cost contributors are identified, one should ensure not to block other potential collaborators to participate in the federated structures. In other words, established cost contributors should not have an exclusive “first mover” advantage. Unless such issues are solved, it is difficult to develop infrastructures that will assure data quality. We thus need to use combinative capabilities in order to divest professional and data quality management to individual resources at the research institutions’ level. In theory this means that a service-oriented access governance structure may be adopted because it requires only limited scientific involvement in administrative tasks. Costs and funds resources will thus be more transparent to the individual characteristics of the projects. In practice however, such decentralised arrangements work best for recent cohorts while relatively mature cohorts still need close collaborations with local scientists. Moreover, outsourcing of data quality management and administrative duties to a task-oriented team requires resources that are not available for many cohorts.