← All guides

The 18 HIPAA identifiers: the Safe Harbor list, PHI and PII examples, and where identifiers hide in documents

HIPAA's Safe Harbor method names 18 identifiers that have to come out before health information counts as de-identified. The list is short, but the identifiers hide in places a form-field checklist misses: fax headers, date stamps, barcodes, handwriting and file metadata.

THE SHORT ANSWER

The 18 HIPAA identifiers are the items HIPAA's Safe Harbor method (45 CFR 164.514(b)(2)) requires removing, for the person and their relatives, employers and household members: names, small geographic areas, dates, contact details, record and account numbers, vehicle and device IDs, URLs, IP addresses, biometrics, full-face photos and other unique codes.

How this page was written. GritWorks makes GritRedact, GritTransform and GritRender, so we are not neutral. What we say about other products reflects their own public websites and documentation as of 21 November 2026. We make no claims about other products' accuracy. Products change: confirm details with each vendor.

The 18 HIPAA identifiers, and where they appear in documents

45 CFR 164.514(b)(2)(i) lists these identifiers, which must be removed for the individual and for their relatives, employers and household members:

#IdentifierWhere it often appears in documents
1NamesPatient name on forms, letterheads, wristband labels, signatures and file names
2Geographic subdivisions smaller than a state: street address, city, county, precinct, ZIP code and their geocodes (the first three ZIP digits may stay if all ZIP codes sharing them hold more than 20,000 people; otherwise they become 000)Address blocks, referral letters, claim forms, map pins
3All elements of dates (except year) directly related to the person, such as birth, admission, discharge and death dates, and all ages over 89 and date elements that indicate such an age (which may be grouped as 90 or older)Date of birth fields, received and fax date stamps, admission and service dates
4Telephone numbersContact fields, fax cover sheets, letterhead footers
5Fax numbersFax headers printed across the top of every received page
6Email addressesContact fields, portal messages, printed email threads
7Social Security numbersRegistration forms, older insurance cards, claims
8Medical record numbersPage headers and footers, barcodes, specimen labels
9Health plan beneficiary numbersMember ID cards, claim forms, eligibility printouts
10Account numbersBilling statements, patient balance letters
11Certificate and license numbersDriver's licence copies, professional licence fields
12Vehicle identifiers and serial numbers, including license platesAccident and injury claim documents
13Device identifiers and serial numbersImplant records, device logs, equipment printouts
14Web URLsPatient portal links, personal pages in correspondence
15IP addressesPortal and telehealth logs, form submission records
16Biometric identifiers, including finger and voice printsFingerprint fields on ID copies, voice recordings
17Full-face photographs and comparable imagesID card copies, clinical photos, video stills
18Any other unique identifying number, characteristic or codeStudy IDs derived from personal data, rare characteristics, unique codes in barcodes and QR codes

Removing identifiers is one route: Safe Harbor and Expert Determination

HIPAA recognizes two ways to show that health information is de-identified and so no longer PHI (45 CFR 164.514(b)).

Safe Harbor. Removing the identifiers is one route, and it has a second condition: the covered entity must also have no actual knowledge that the remaining information could identify the person, alone or in combination with other information. A rare diagnosis, an unusual job or a detail in a free-text note can still point to someone after all 18 identifiers are gone. A covered entity may keep a re-identification code of its own if the code is not derived from information about the person and the way to decode it is not disclosed.

Expert Determination is the other route. A person with appropriate statistical and scientific knowledge determines that the risk that the anticipated recipient could identify anyone is very small, and documents the methods and results. It can keep data that Safe Harbor removes, such as dates or finer geography, when the analysis supports it.

Redaction alone is not HIPAA de-identification. Removing values from a document is a technique. De-identification is an outcome that one of the two methods has to establish, with policies and validation that fit the intended use and recipient. For the wider difference between PHI and PII, see the PHI vs PII guide; for anonymisation outside HIPAA, the digitally anonymised guide.

PHI examples and PII examples

PII examples (information that identifies a person anywhere): name, home address, date of birth, phone number, email address, Social Security number, passport or driver's licence number, bank and card account numbers, IP address, fingerprints, voice recordings and photos of a face.

PHI examples (identifiable health, care or payment information held by a HIPAA covered entity or business associate): a lab report with the patient's name, a claim form with a member ID and diagnosis codes, a discharge summary, an appointment schedule, a prescription record, a hospital bill, a radiology image with the medical record number in its header.

ExamplePHI, PII or neither?
Patient name and date of birth on a lab report held by the laboratoryPHI, and also PII
Member ID and diagnosis codes on a claim held by a health planPHI, and also PII
Social Security number on a loan application held by a bankPII, not PHI: a bank is not a covered entity for that record
A sick note in the HR file of a hospital acting as employerPII; excluded from PHI as an employment record
Heart rate in a consumer fitness app that is not a covered entity or business associateUsually PII outside HIPAA; other laws may apply
Diagnosis counts by state with all 18 identifiers removed and no actual knowledge of identifiabilityNeither: de-identified under Safe Harbor

Where identifiers hide in documents

  • Headers and footers. Fax headers print the sender's number, a date and time and sometimes a name on every page. Many records repeat the patient name and record number in each page footer.
  • Stamps. "Received" and "Reviewed" stamps add dates, which are identifiers when they relate to the person's care.
  • Handwriting. Margin notes, signatures, initials and corrections are missed by tools that read only typed text.
  • Barcodes and QR codes. Specimen labels, wristbands and forms encode record numbers and names that are invisible to text search.
  • Images. Copies of ID and insurance cards carry faces, numbers and addresses; clinical photos can show faces or distinctive features.
  • Metadata and file names. PDF author and title fields, comments, earlier revisions, image EXIF data and file names such as a patient's surname with a date.
  • Free text. Narrative notes mention relatives, employers, places and events that identify the person, which falls under the catch-all identifier and the actual-knowledge condition.

Removing a value means removing it from every layer: the text, the page image, any hidden OCR text and the metadata. The what is redaction guide and the PII redaction guide cover how to do that and how to check it.

CMS-1500 claim: readable sample from the Insurance pack

FREE SAMPLE PACK · INSURANCE

Try it with the free Insurance sample pack.

Fictional documents, clean and degraded, with the expected value and position of selected populated fields.

10 PDFs · Premium statement, CMS-1500 claim, Itemized invoice and more · JSON and bounding boxes · CC BY 4.0

Free sample · No call required · No platform account

Need larger volumes, your own layouts or other formats? Talk to us ↗

Testing redaction without real PHI

To test OCR, extraction or redaction against the 18 identifiers, use fictional documents whose values are known in advance, then search the output for every one of them. The free insurance sample pack follows one fictional health-insurance case through claims and statements; it is a testing resource, not a free tier. GritRender generates larger sets with known values.

GritRedact detects and permanently removes PII and PHI from documents, images and audio, with human review before removal and an audit record. Its output still needs a de-identification method and a process that match the intended use.

Not legal advice. This is an educational summary prepared by GritWorks, a software company, not a law firm. Whether a record is PHI, and whether a dataset is de-identified, depends on your facts; confirm with your privacy officer or counsel.

What GritWorks does not replace

  • GritWorks complements authorization, governance and e-discovery tools rather than replacing them. Access control, data catalogues, retention policy and legal review platforms keep their jobs; GritWorks prepares the documents, images and audio that move between them.
  • Redaction is not authorization. If a person or system is not entitled to a record, block access instead of sending a redacted copy.
  • Automated redaction alone is not a declaration of HIPAA de-identification. HIPAA recognizes the Safe Harbor and Expert Determination methods; policies, validation and human review must match the intended use and recipient.
  • PII flagging is not redaction. GritTransform can flag potentially sensitive values in extracted results, but that does not remove them from the source. Use GritRedact when content must be removed.
  • Engagement is sales-assisted. GritWorks has no public or self-serve API, SDK, MCP server or free tier, and publishes no price list: pricing is agreed with sales. Customer APIs are provided after onboarding.

Common questions

What are the 18 HIPAA identifiers?

Names; geography smaller than a state; date elements other than the year, and ages over 89; phone and fax numbers; email addresses; Social Security, medical record, health plan beneficiary, account and certificate or license numbers; vehicle and device identifiers; web URLs; IP addresses; biometrics; full-face photos; and any other unique identifying number, characteristic or code.

What are examples of PHI?

A lab report with the patient's name, a claim form with a member ID and diagnosis codes, a discharge summary, a prescription record or a hospital bill, when held by a HIPAA covered entity or business associate. Health information becomes PHI when it identifies the person.

What are examples of PII?

Names, home addresses, dates of birth, phone numbers, email addresses, Social Security numbers, passport and driver's licence numbers, bank and card account numbers, IP addresses, fingerprints, voice recordings and photos of a face: anything that identifies a person alone or in combination.

Is removing the 18 identifiers enough to de-identify data?

Removing them is one route, Safe Harbor, and it also requires that the covered entity has no actual knowledge that what remains could identify the person. Expert Determination is the other route. Redaction alone is not HIPAA de-identification.

Are dates HIPAA identifiers?

Yes. All elements of dates except the year are identifiers when they relate directly to the person, such as birth, admission, discharge and death dates. Ages over 89 are identifiers too and may be grouped as 90 or older.

Is a ZIP code a HIPAA identifier?

Yes. Under Safe Harbor only the first three digits may stay, and only if all ZIP codes sharing them hold more than 20,000 people according to current Census data; otherwise those digits are changed to 000.

Is this page legal advice?

No. It is an educational summary prepared by GritWorks, a software company, not a law firm. Confirm how HIPAA applies to your records with your privacy officer or counsel.

CMS-1500 claim: readable sample from the Insurance pack

FREE SAMPLE PACK · INSURANCE

Try it with the free Insurance sample pack.

Fictional documents, clean and degraded, with the expected value and position of selected populated fields.

10 PDFs · Premium statement, CMS-1500 claim, Itemized invoice and more · JSON and bounding boxes · CC BY 4.0

Free sample · No call required · No platform account

Need larger volumes, your own layouts or other formats? Talk to us ↗

A USEFUL START, IN YOUR INBOX

Your next dataset.
Delivered.

10 PDFs. Clean and challenging versions. JSON and bounding boxes included.

Use the inbox where you’d like to receive your pack.

Free sample · No call required · No platform account

Start with your workflow

Put your unstructured data to work.

Tell us what your team needs to achieve. We’ll help you choose the right products, work through your requirements and explore on-premises or SaaS deployment.

Request a working session