Glossary
De-identification
Removing or transforming personal identifiers in health data so that individuals cannot reasonably be identified, enabling research use.
De-identification is the process of stripping direct identifiers (names, health-card numbers, addresses, exact dates) and reducing indirect identifiers (rare combinations of age, location and diagnosis) so that a dataset can be used for analysis and research without exposing the people in it. Techniques range from removing fields and generalizing dates to statistical methods that measure re-identification risk.
In Canada, provincial health-privacy laws such as Ontario's PHIPA and the federal PIPEDA set the expectations for how personal health information is handled; de-identification is what makes "research-ready" data possible while meeting those expectations. It is not a one-time step but an ongoing practice, with governance over who can access what and for which purpose.
Why it matters to our work
We build de-identified, governed data foundations from the start so that care data can support research and audit without compromising patients' privacy.