PHI Patterns
Scrub detects the following Protected Health Information patterns.
Identity Patterns
Social Security Number (SSN)
Detects US Social Security Numbers in various formats.
Examples detected:
123-45-6789123 45 6789123456789(when context suggests SSN)
Date of Birth (DOB)
Detects birth dates in common formats.
Examples detected:
01/15/1985January 15, 19851985-01-15DOB: 01/15/85
Phone Numbers
Detects US phone numbers in various formats.
Examples detected:
(555) 123-4567555-123-4567555.123.4567+1 555 123 4567
Email Addresses
Detects email addresses.
Examples detected:
patient@example.comjohn.doe@hospital.org
Medical Identifiers
Medical Record Number (MRN)
Detects medical record numbers with common prefixes.
Examples detected:
MRN: 12345678MRN#12345678Medical Record: 12345678
Medicare ID
Detects Medicare Beneficiary Identifiers (MBI).
Examples detected:
1EG4-TE5-MK72Medicare: 1EG4TE5MK72
Medicaid ID
Detects Medicaid identification numbers.
Examples detected:
Medicaid: 123456789Medicaid ID: AB12345678
Health Plan ID
Detects health insurance plan identifiers.
Examples detected:
Policy: XYZ123456789Member ID: 123456789
Location Patterns
Street Addresses
Detects physical addresses including street, city, state, and ZIP.
Examples detected:
123 Main Street, Boston, MA 02101456 Oak Ave, Apt 2B789 Hospital Drive
PO Box
Detects PO Box addresses.
Examples detected:
PO Box 12345P.O. Box 12345Post Office Box 12345
ZIP Codes
Detects US ZIP codes (more than 3 digits).
Examples detected:
0210102101-1234
Other Identifiers
Driver's License
Detects driver's license numbers with state prefixes.
Examples detected:
DL: S12345678License: 123456789
Vehicle Identifiers
Detects VINs and license plate numbers.
Examples detected:
VIN: 1HGBH41JXMN109186Plate: ABC-1234
Account Numbers
Detects various account number formats.
Examples detected:
Account: 123456789Acct# 123456789
Device Identifiers
Detects medical device serial numbers and UDIs.
Examples detected:
Device SN: ABC123456UDI: (01)12345678901234
Biometric Data
Flags mentions of biometric identifiers.
Examples detected:
- References to fingerprints, retinal scans, voiceprints
Photo/Image References
Flags references to identifying photographs.
Examples detected:
[Photo attached]See patient image
Pattern Accuracy
Scrub uses a combination of:
- Regular expressions for structured data (SSN, phone, email)
- Context analysis for ambiguous patterns (names, dates)
- Format validation for IDs (Medicare, MRN)
This approach minimizes false positives while catching real PHI.
Limitations
Some PHI types are difficult to detect automatically:
- Common names without context (e.g., "John" alone)
- Dates that could be PHI or general dates
- Free-text descriptions of identifying characteristics
For maximum protection, consider using Block or Redact mode and training staff on PHI awareness.