The file arrived with no warning. Inside, rows of names, medical records, and birth dates—clear, unprotected, waiting to be copied. This is PHI sensitive data, the kind that triggers breach reports, audits, and fines. It is also the kind of data you cannot afford to mishandle.
PHI, or Protected Health Information, includes any data that can identify a patient when combined with health or medical details. It is regulated under HIPAA in the United States. PHI sensitive data goes beyond just obvious identifiers like Social Security Numbers. A phone number tied to a lab result is PHI. An email linked to a diagnosis is PHI. Even metadata that can be traced back to a patient can count.
Handling PHI sensitive data requires controls at every stage—collection, storage, transmission, and disposal. At rest, data should be encrypted with strong keys. In transit, use secure protocols like TLS 1.2 or higher. Limit access using role-based permissions, and log every query, update, or export event. Audit these logs often.
Developers should design systems so PHI sensitive data is isolated from other datasets, with separate storage and strict API-level filtering. Mask or tokenize identifiers when use of real data is not required. Never store PHI in code repositories, analytics dashboards, or logs. If possible, use synthetic data during development and testing.