What is meant by de-identified data in the context of patient information disclosures?

• A. Identifiers removed so individuals cannot be readily identified
• B. Data remains fully identifiable but encrypted
• C. Data is stored in a public repository
• D. Data includes full identifiers to maximize accuracy

Answer: (D)

De-identification means removing or masking the personal identifiers that could reveal who a person is, so the data cannot be readily linked back to an individual. This typically involves taking out names, addresses, phone numbers, dates of birth, medical record numbers, social security numbers, and other direct or indirect identifiers. In privacy practice, there are formal methods such as Safe Harbor (removing a defined set of 18 identifiers) or Expert Determination (a professional assesses and certifies that re-identification risk is very low).

So the key idea is to strip out or mask identifiers to protect privacy, not keep them or rely on encryption alone. Data that still includes full identifiers would not be considered de-identified, and encryption, while protective, doesn’t remove identifiers from the dataset. Public repositories can host de-identified data, but de-identification itself is about the removal of identifiers, not the data’s storage location. Also, even de-identified data carries some residual re-identification risk if combined with other information, so appropriate safeguards and governance remain important.