Health & Life Sciences

Direct identifiers such as names, medical record numbers and insurance identifiers, and indirect identifiers such as dates of service, postal codes and rare diagnoses that can single out an individual in combination. Both categories need treatment before data leaves a clinical system. Mage Data ships a HIPAA classification policy that covers both out of the box, with each identifier mapped to a fit-for-purpose protection method.

Direct identifiers are removed or replaced with surrogates, and quasi-identifiers such as dates and geography are generalized so that individuals cannot be singled out. The technique depends on whether the recipient needs to follow a patient over time. Mage Data offers the full range: suppression, generalization, date shifting, consistent surrogates and Identities, which replaces real patients with realistic fictitious ones, plus synthetic data generation.

Yes. Mage Data maps the same patient identifier to the same surrogate value across every table, system and extract, and shifts dates consistently for each patient, so the care timeline stays intact and analyzable without the real identifier appearing anywhere in the dataset.

Identifiers in free text are not confined to a labeled column, so column-name rules miss them. Mage Data uses natural language processing and large language models to detect names, dates, locations and identifiers inside notes, letters, reports and scanned documents, so they can be masked alongside structured fields under the same policy.