The iHi-funded research project, iCARE4CVD, is a collaborative effort to personalize prevention and treatment of cardiovascular disease (CVD), and has just taken a big step forward in achieving this goal.
Made up of 10 different teams called ‘Work Packages’ (WPs), the project splits up its ambitious goals into smaller pieces, each with a different focus. One such team, WP2, focuses on producing a type of database management system – a federated database – that combines multiple independent databases into a transparent and centralized system. The goal of this database is to provide a central hub where diverse bundles of high-quality data can be collected, protected, and analysed.
Co-led by Maastricht University and Decentriq, WP2 has achieved a significant milestone in the project by building the first version of its federated database, comprised of a cohort explorer and data clean rooms. This cutting-edge platform allows researchers to share and analyse any health data collected by iCARE4CVD safely.
The Cohort Explorer
A cohort within iCARE4CVD refers to a group of patients who all share a similar characteristic binding them together, such as the type of diagnostic screening, comorbidity, wearable data, or biomarker. Developed by the University of Maastricht, the cohort explorer, as its name suggests, will allow partners within the project to explore the descriptive information of cohorts to better organize and understand them. Unique to iCARE4CVD, and facilitated by Decentriq, users will be able to conduct research on cohorts that have been combined and analyse them at the patient level – a stark contrast to traditional collaboration setups that favour data combined from multiple sources and then summarized.
Decentriq’s Data Clean Rooms
An important feature of the technology powering Decentriq is that it not only allows users to analyse patient data, but it allows them to go into specifics without revealing any private information. Using confidential computing as a basis, the data clean rooms allow participants to analyse data while it remains encrypted. Additional privacy-enhancing technologies and control mechanisms built into the data clean rooms allow data custodians (those responsible for the management and protection of data) to control exactly how their data is used beforehand, while assuring that analysts can only extract privacy-preserving results of models and computations — never raw data itself. These measures guarantee that data analysts can perform their usual tasks compliantly, without the risk of leaking sensitive patient information from the data clean room.
Future Outlook
Decentriq has already helped onboard two data partners (Uniklinik Aachen and University of Maastricht) onto the federated database during the third month of the project, speeding up the collaboration process. As iCARE4CVD looks to bring in more data and more cohorts on cardiovascular patients, WP2 will facilitate the onboarding of this information as well. With a federated database that will one day contain diverse high-quality data, iCARE4CVD will perform analysis to diagnose patients with CVD earlier, define their risk of adverse health outcomes, and predict their response to treatment.
