Document de-identification for research and outsourcing
Short answer: De-identification removes information that directly identifies a person before documents go to a research, analysis or outsourcing team. EnjoyPDF processes scans in batches on your own computer, and the operation record contains only categories and positions, not the deleted content.
Why this use case needs offline processing
Research data, medical records and customer files often come with explicit requirements that data must not leave the organization. Uploading files to an online tool for redaction would itself break that requirement. With PDF Text Replacer, recognition, deletion and rewriting all happen on your own computer.
Batch workflow
- In the main window, switch to “Secure redaction”, tick the ID types, and add custom keywords such as names and medical record numbers (you can import them from the first column of a spreadsheet).
- Open the batch window, add the folder, and choose “Secure redaction” as the operation.
- Each output file has its own residue check result, and the processing report summarizes all files.
- Give the output files and the processing report to the recipient. The operation record (manifest.json) does not contain the deleted content, so it can be delivered with them.
What you need to add
The built-in rules cover ID numbers, bank card numbers, mobile numbers, international phone numbers and email addresses. Add the following as custom keywords to suit your project:
- Names, including names of relatives.
- Medical record numbers, study subject IDs, employee numbers.
- Specific addresses and organization names.
- Dates directly related to an individual.
Related regulations
The US HHS gives two methods for HIPAA de-identification: Expert Determination and Safe Harbor. China’s Personal Information Protection Law and Thailand’s PDPA also define de-identification and anonymization. This page is not legal advice.
FAQ
Does it meet the HIPAA de-identification standard?
The software removes the identifiers you specify and verifies that they are fully removed. The HIPAA Safe Harbor method requires removing 18 categories of identifiers, many of which (such as dates and address details) you need to add as custom keywords. Whether the standard is met should be decided by your compliance officer.
Can it be used on a computer without internet access?
Yes. Document processing does not use the network, and the current version has no automatic network activity. Language packs can be downloaded on a connected computer and imported.