---
title: "PHI Redaction for Healthcare"
description: "Redact PHI on a Windows PC with Philter Desktop, or in your own cloud with Philter. For hospitals, health systems, and medical and dental practices."
url: "https://philterd.ai/use-cases/healthcare/"
---

[← All industries](https://philterd.ai/use-cases/)

Hospitals and medical practices

# PHI Redaction for Healthcare

Redact protected health information from referrals and records on a single PC, or from clinical data flowing out of your EHR. HIPAA treats data de-identified by its Safe Harbor or Expert Determination methods as outside the Privacy Rule, which is why teams redact before data reaches analytics, research, or AI.

![A referral letter from a family medicine practice to a cardiologist, listing the patient's name, date of birth, address, medical record number, phone, and email, and describing her symptoms](https://philterd.ai/images/use-cases/healthcare-letter-original.png)

Original

![The same referral letter after Philter's HIPAA Safe Harbor policy, with the dates, names, street address, city, medical record number, phone, email, and age blacked out, and the ZIP code cut to 506\*\*. The clinical description is left readable.](https://philterd.ai/images/use-cases/healthcare-letter-redacted.png)

Redacted

Invented data, redacted by Philter 4.0.0 with the library's HIPAA Safe Harbor policy. Detection is configurable and probabilistic, so review redacted output.

For hospitals and health systems

[Philter, in your own cloud](https://philterd.ai/philter/)

[Deploy](https://philterd.ai/philter/deploy/)

- [AWS Marketplace](https://aws.amazon.com/marketplace/pp/B07YVB8FFT?ref=_ptnr_mf_launch)
- [Azure Marketplace](https://azuremarketplace.microsoft.com/en-us/marketplace/apps/philterdllc1687189098111.philter?tab=Overview)
- [Google Cloud Marketplace](https://console.cloud.google.com/marketplace/product/philterd-public/philter)

For clinics and medical or dental practices

[Philter Desktop, on your own computer](https://philterd.ai/philter-desktop/)

[Download](https://philterd.ai/philter-desktop/download/)

- Windows PC

## What Philterd does with patient data

### The same patient, the same stand-in

Philter can replace a value with the same stand-in everywhere it appears, so records about one patient still link after redaction. If you keep a way back to the original, 45 CFR 164.514(c) sets the conditions for that re-identification code.

### Starts from HIPAA Safe Harbor

Philter Desktop includes a Safe Harbor template, and the policy library has one for Philter. Both are starting points. Identifiers in your own formats, such as medical record and health plan numbers, need patterns you add, and Safe Harbor also requires that you have no actual knowledge the remaining data could identify someone.

### A person reviews the result

Philter Desktop shows each redaction before saving and can report every change. For pipelines, [Philter Workspace](https://philterd.ai/philter-workspace/) gives reviewers a place to accept or override what Philter found, with each decision recorded.

## Which fits your organization

A practice, one PC at a time

### [Philter Desktop](https://philterd.ai/philter-desktop/)

Redact referrals, records, and forms on a Windows PC, including scanned faxes saved as PDFs. Patient documents stay on that computer.

[About Philter Desktop →](https://philterd.ai/philter-desktop/)

A practice sharing the work

### [Philter One](https://philterd.ai/philter-one/)

Coming in the first half of 2027. A small computer on the practice network that front desk and clinical staff use from a web browser, with a person’s review before every redacted copy.

[Join the list →](https://philterd.ai/philter-one/)

A hospital or health system

### [Philter](https://philterd.ai/philter/)

Redacts clinical notes, EHR exports, and documents through an API in your own cloud. It runs in production in an [EHR data pipeline](https://philterd.ai/case-studies/#ehr-data-pipeline) on AWS. Philterd never receives patient data, so there is no BAA to sign with us.

[About Philter →](https://philterd.ai/philter/)

A team building a healthcare app

### [PHI Assessment](https://philterd.ai/consulting/healthcare-app-phi-assessment/)

A fixed-price, read-only review of where patient data goes in your app, with a location for every finding and a plan you can show a reviewer.

[See the assessment →](https://philterd.ai/consulting/healthcare-app-phi-assessment/)

## Getting started in a hospital or health system

1. **Deploy Philter from your cloud’s marketplace** into the same cloud account as your EHR exports, so the clinical data stays in one account.
2. **Start from the HIPAA Safe Harbor policy** in the policy library, and add patterns for the identifiers your EHR uses, such as medical record numbers.
3. **Measure it on your own records.** Have a clinical reviewer mark a sample, then score the policy against it with [Philter Scope](https://philterd.ai/philter-scope/). Precision and recall like this is the kind of evidence an Expert Determination relies on, though the determination itself is the expert’s.
4. **Connect it to one pipeline at a time,** such as an EHR export, claims processing, or the documents behind an AI feature.

[Deploy in your cloud](https://philterd.ai/philter/deploy/) [Plan a pipeline with us](https://philterd.ai/consulting/pii-data-pipelines/)

## Ready-to-use redaction policies

Free, ready-to-use policies from the [open source policy library](https://philterd.ai/policies/). Download and load into your Philter instance.

Healthcare v2.0.1

### [HIPAA Safe Harbor De-Identification](https://philterd.ai/policies/healthcare/hipaa-safe-harbor/)

A starting point for HIPAA Safe Harbor de-identification of clinical text per 45 CFR 164.514(b)(2).

HIPAA Safe Harbor PHI 45 CFR 164.514

Healthcare v2.0.0

### [Clinical Notes De-Identification (Date-Shifted)](https://philterd.ai/policies/healthcare/clinical-notes-deid/)

De-identify clinical notes for research, ML training, or analytics — preserving temporal relationships via per-patient date shifting.

HIPAA PHI clinical notes date shifting

Healthcare v2.0.0

### [Medical Chatbot — User Input Redaction](https://philterd.ai/policies/healthcare/medical-chatbot/)

Redact PHI from user messages to a healthcare chatbot before they reach the LLM — preserves clinical meaning while removing identifiers.

HIPAA PHI chatbot LLM

[Browse all redaction policies →](https://philterd.ai/policies/)

## Recent writing on healthcare

### [Automating HIPAA Safe Harbor: A Blueprint for Healthcare Data Pipelines](https://philterd.ai/blog/automating-hipaa-safe-harbor-a-blueprint-for-healthcare-data-pipelines/)

How the Philterd suite maps to the 18 HIPAA Safe Harbor identifiers, with a deployment blueprint for patient data lakes, research pipelines, and medical RAG.

### [Five Ways a Healthcare App Leaks PHI Without Anyone Noticing](https://philterd.ai/blog/five-ways-a-healthcare-app-leaks-phi/)

The demo looks clean. Here are five places PHI quietly ends up in a fast-built healthcare app, why each one happens, and what to do about it.

### [Redact PDFs Offline on Windows so No Data Leaves Your PC](https://philterd.ai/blog/redact-pdfs-offline-on-windows/)

Philter Desktop redacts PDFs entirely on your Windows PC, including scanned pages read with on-device OCR. No uploads, no server, and it works offline.

[All blog posts →](https://philterd.ai/blog/)
