Title of Presentation
“Linking data without privacy breaches”
Date and Place
Professor Ronald Stolk is an internationally recognized expert in biobank/cohort studies and related Big Data in Health. My current position is director “Research Data & Biobanking (RD&B)” within the Office of the Dean of Research, University Medical Center Groningen. With the RD&B team we make policies on sample storage, data stewardship, research quality, etc. Together with the board of the university computer center we work on innovative IT structures for human subject research. I am the founding director of BiKE (Biobank/cohort Knowledge and Expertise center).
To support biobank and data-based research, I am member of national and international organizations on research infrastructure and Big Data in Health. Among others: BBMRI-NL, DTL/ELIXIR-NL, BBMRI-ERIC, Maelstrom Research.
I thrive on connecting IT experts, clinical researchers and health care professionals to foster the use of Big Data in Health. I am keen on privacy and security of these data.
Biobank studies increasingly moves into databank research. Contemporary research questions increasingly include data that is not collected within the biobank study, notably data from clinical practice: Big Data in Health. Linking data on an individual level involves potential privacy risks, especially since health data are considered sensitive data. The recently approved European General Data Protection Regulation (GDPR, Regulation 2016/679) requires “data protection by design”. This means that privacy-friendly techniques such as pseudonimisation and encryption will be encouraged. When linking data additional privacy risks arise, notably the potential for re-identification. Also the informed consent procedure and feedback to patients/participants is challenging with data from multiple sources.
Data linkage through a Trusted Third Party (TTP) is the most commonly used technique to prevent privacy breaches. Promising new technologies include blockchain (the technology used for bitcoin) and Linked Data.