Building a synthetic data network in Canada

Building a Synthetic Data Network in Canada

Using privacy preserving synthetic data to connect hospitals across Canada and unlock faster insights to improve patient care

Hospital data within the Canadian healthcare landscape remains siloed and underutilized due to strict privacy rules and lack of cross institutional data sharing. At the same time, the Canadian health system is under pressure from rising costs, capacity challenges, aging populations and a fragmented research and innovation ecosystem. This multi-site proof of concept project addresses that challenge by setting the foundation for a scalable, pan-Canadian synthetic data network that allows hospitals to collaborate without sharing sensitive patient information. Synthetic data is non-reversible and artificially created synthesis of real patient data. The dataset replicates the statistical characteristics and correlation of the raw data. Synthetic data use enables safe, compliant data access across hospitals and provinces.

The project will initially focus on non-small cell and small cell lung cancer (NSCLC) leveraging synthetic data from McGill University Health Centre (MUHC) and University Health Network (UHN). Raw hospital data will be standardized and used to generate a synthetic dataset that preserves statistical characteristics in the data while protecting personal health information. Using advanced analytics, we will analyze treatment pathways, patient outcomes, and healthcare resource use. A federated learning approach will also be used, allowing models to learn from synthetic data and maintain patient privacy.

This initiative has the potential to unlock the power of data to drive efficiency, improve patient care and accelerate innovation within Canada. This proof of concept aims to demonstrate that synthetic data can enable secure, cross-institutional collaboration and generate reliable insights. These insights could improve treatment decisions, identify gaps in care, and support more efficient and equitable healthcare delivery. In the long term, this work will lay the foundation for a scalable, privacy-preserving national data network to accelerate precision medicine research and improve patient outcomes across Canada.

“Successfully creating high-quality synthetic data will allow us to apply data science research in ways that does not compromise either patient privacy or validity of the findings. This opens a new door for precision medicine, enabling discoveries that may not have otherwise been possible.” - UNB Research Team