Decentriq to facilitate analysis of data from over one million cardiovascular disease patients

About iCARE4CVD

Cardiovascular diseases (CVDs) are very prevalent worldwide, with enormous socio-economic impact. They are still the most common cause of death, despite significant advancements in therapy. Due to population aging and unhealthy lifestyles, the European prevalence of CVD (>85 million) is on the rise, underscoring the critical need for better care pathways to reduce the impact of CVDs.

“iCARE4CVD”, an international, public-private research consortium launched on October 25, 2023 is funded by the Innovative Health Initiative (IHI) — a joint undertaking of the European Commission and the European life science industry.

The consortium is one of the first to be funded by the new IHI framework (former IMI) that is funded by both industry and public partners, comprises 33 prominent global collaborators from various sectors, and will run for 54 months. Using a combination of artificial intelligence, biomarkers, and large datasets, it seeks to personalize treatments and improve outcomes of patients affected by CVDs by enabling:

• Enhanced and timelier disease diagnosis
• Tailored treatment options rooted in patient sub-phenotyping and risk assessment
• Augmented comprehension of disease progression and non-responsive cases
• Discovery of novel healthcare pathways
• Enhanced cost-effectiveness through the delivery of appropriate treatment to the correct patient at the appropriate moment

The role of patient data in improving CVD outcomes

Spanning 12 countries, the consortium includes 10 life sciences companies who will work together with 13 hospitals, academic centers and research organizations to analyze the data of over one million CVD patients. The results of the analysis will help researchers better understand what factors contribute to the development of the diseases and how to best adjust patient treatment accordingly.

It is crucial that all data is processed and analyzed in a way that preserves patient privacy and complies with regulatory requirements surrounding patient data.

How Decentriq ensures patient data privacy and enforces compliance

Decentriq’s data collaboration platform is a core component of iCARE4CVD’s data architecture. In the platform, participants contribute their data for research in data clean rooms, secure environments protected by a combination of advanced privacy technologies including confidential computing. With this unique approach, raw data remains on-premise and completely inaccessible to any other participant, even while analysis is underway — facilitating in-depth research while enforcing data privacy and compliance. 

Decentriq performs three essential functions for the consortium’s research: facilitating data encryption, preserving the privacy of any results generated, and machine learning analysis. 

Facilitating end-to-end data encryption 

Decentriq establishes a secure infrastructure for analysis that encrypts the data even during computation. Using confidential computing as a basis, the data clean rooms allow participants to analyze data while it remains encrypted. The clean room additionally provides verification via logs of what has happened to the datasets as well as proof that raw data remained encrypted through the entire process. 

Preserving the privacy of results generated

Additional privacy-enhancing technologies and control mechanisms built into the data clean rooms assure 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 inadvertently leaking sensitive information from the data clean room environment.

Enabling machine learning analysis

Researchers will develop machine learning models within the data clean rooms. These will function as a comprehensive analysis environment similar to centralized machine learning platforms. Analysts can seamlessly write and execute their analyses using familiar tools like Python, R, and SQL. Synthetic data allows analysts to explore possibilities for analysis without accessing raw data. Crucially, this all happens within the encrypted confines of the clean room, guaranteeing the security of the data and the control of the output obtained by the analysis.

Empowering sustainable CVD research through data collaboration

The collaboration made possible through Decentriq’s data clean rooms supports data-driven advancements to CVD diagnosis and treatment in a sustainable, ethical manner.

To ensure sustainable availability of the datasets used, the consortium will adhere to FAIR principles (Findability, Accessibility, Interoperability, and Reuse). The datasets can be found with the help of detailed metadata and persistent IDs, accessibility ensured with a stable repository, interoperability promoted through standardized formats, and reusability enhanced with comprehensive documentation. This approach fosters collaboration and innovation while maintaining patient data confidentiality.

In addition to improving patient outcomes, the use of data clean rooms will foster sustainability in CVD research by optimizing resource usage, reducing unnecessary medical interventions, safeguarding data privacy, and promoting ethical data handling.

Learn about more about how secure data collaboration facilitates real-world data use in the white paper: Unlock the value of real-world healthcare data with confidential data clean rooms.

Funding acknowledgement

This project is supported by the Innovative Health Initiative Joint Undertaking (IHI JU) under grant agreement No 101112022. The JU receives support from the European Union’s Horizon Europe research and innovation programme and COCIR, EFPIA, Vaccines Europe, EuropaBio, MedTech Europe and JDRF International.

Disclaimer

Funded by the European Union, the private members, and those contributing partners of the IHI JU. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the aforementioned parties. Neither of the aforementioned parties can be held responsible for them.