opinion: canada must own the systems that connect its health data

every day across canada, enormous quantities of health data are generated: medical images, pathology reports, genome sequences, treatment decisions and patient outcomes. collectively, these data hold valuable clues that could help us detect disease earlier, understand why one treatment works for one patient but not another, and develop entirely novel approaches to care.

yet much of this data remains fragmented across our many institutions and jurisdictions. the result is a paradox; canada has the raw materials—the data—to unlock discoveries that will improve the health of our citizens, but lacks the infrastructure needed to safely and securely realize that promise.

this challenge is becoming increasingly important as health care moves toward more personalized approaches to diagnosis and treatment. precision medicine depends on finding patterns across clinical records, detecting molecular changes, genomics, imaging and more. artificial intelligence can help reveal those patterns, but only if researchers can access large and diverse datasets.

clearly, the management of health data carries unique responsibilities. information is governed locally, protected for privacy, and stewarded by institutions that have gained the trust of patients and communities. any effort to expand access to health data must respect those necessary obligations.

the solution is to keep data where it is already most secure: inside the health-care facilities that generated it while under privacy regimes established to protect against misuse. what has to change is our ability to query that data across many sites at once.

a federated learning platform offers that model. patient-level data remain with the hospital or institution that collected them, behind local firewalls and under local governance. approved analytical queries or tasks are distributed to participating sites, enabling analyses across datasets that, cumulatively, are large enough to reveal new therapeutic avenues for rare diseases, for example, without accidentally disclosing individual patient information.

in other words, computation is what moves, not the raw records.

this computational architecture lets researchers tackle complex health challenges by learning from diverse datasets across organizations and provinces without centralizing or transferring sensitive information between jurisdictions, enabling insights that no single institution could generate on its own.

but technology alone isn’t enough.

a national health data ecosystem cannot be built by a single institution. success depends on collaborating with the people who generate, govern, use and are represented in the data.

researchers generate scientific questions and real-world workflows. health-care institutions understand clinical operations, stewardship and accountability. patients bring essential perspectives on consent, transparency and trust. indeed, patients expect their data is being used appropriately to improve outcomes for others. privacy and cybersecurity experts help ensure proper safeguards are in place. canadian companies contribute technical expertise and innovation. each group has a role to play in building solutions that are both effective and trusted.

a spirit of collaboration is essential to developing infrastructure that works in practice, and there are some platforms that have already begun bringing together canadian enterprises, health data providers, and researchers to address real-world challenges and ensure these systems meet the needs of those who will ultimately use them, such as the the digital health & discovery platform being developed by the terry fox research institute.

this is also an opportunity for canada to strengthen its capacity in an area that’s becoming increasingly important globally. health data generated in canada should be governed according to canadian laws and expectations. and building domestic expertise and infrastructure keeps accountability close to patients and creates opportunities for researchers and companies to contribute to the next generation of health innovation.

canada generates large amounts of valuable health data every day. the challenge is ensuring it can be used responsibly, securely and at the scale modern research requires for meaningful outcomes. federated approaches offer a promising path forward, helping to unlock the next generation of innovation rather than keeping this potential for progress trapped within silos.

if we can enable safe, secure interconnection of data analysis and build a trusted foundation for future discovery, canada will not only accelerate research and strengthen innovation, but also expedite better health outcomes for patients across the country.

dr. jim woodgett is the president and scientific director with terry fox research institute, and josip maras is national director, digital health & discovery platform.