---
title: "PHI Sensitive Data Definition (SDD) Example"
url: "https://dm0bkdes5q7a2ls.productfruits.help/en/article/phi-sensitive-data-definition-example"
language: "en"
last_updated: "2026-06-03T16:27:17.796Z"
---

# PHI Sensitive Data Definition (SDD) Example

To create a high-accuracy PHI (Protected Health Information) SDD, Spirion recommends you combine a Personal Identifier (who the person is) with a Medical Context (what their health status is).

Here is a practical example of PHI SDD logic designed to reduce false positives:

### SDD Name: `High-Confidence PHI (Identifier + Clinical Context)`

#### **The Logic Components:**

1.  **Primary Detector:** `AnyFind: Social Security Number` (or `AnyFind: Name`)
2.  **Operator:** `Near with Distance (100)`
3.  **Context Detector:** `Custom Dictionary: Clinical_Keywords` (containing terms like *Diagnosis, ICD-10, Patient, Treatment, Prognosis, Clinic*)

#### **The Logic String:**

> `[AnyFind: Social Security Number] NEAR WITH DISTANCE (100) [Custom Dictionary: Clinical_Keywords]`

* * *

### Why this logic is effective:

-   **Proximity (The "100" Rule):** By setting a distance of 100 characters, you ensure the SSN and the medical term are logically related (e.g., in the same paragraph, form field, or spreadsheet row). An SSN on page 1 and the word "Patient" on page 10 will not trigger this rule.
-   **Validation:** Because it uses the `AnyFind: Social Security Number`, Spirion’s engine automatically performs checksum validation. This means a random 9-digit part number will not trigger the SDD, even if it's near a medical term.
-   **Reduced Noise:** A list of employee names and SSNs for payroll (which is PII, but not PHI) will **not** trigger this SDD because the clinical keywords are missing.

* * *

### Advanced Variation: The "Triple-Match" PHI Rule

If you want to be even more specific (for example, for a "Restricted" classification), you can require 3 elements:

> `([AnyFind: Name] NEAR WITH DISTANCE (50) [AnyFind: Date of Birth]) REQUIRE [AnyFind: Health Information]`

-   **How it works:** This requires a Name and DOB to be right next to each other (establishing a "Person") AND requires the built-in `Health Information` AnyFind to be present anywhere in the same file.

### Implementation Steps

1.  **Create the Dictionary:** Go to `Data Types` -> `Custom` -> `Dictionary` and add your clinical terms.
2.  **Build the SDD:** Go to `Data Types` -> `Sensitive Data Definitions` -> `New`.
3.  **Map to Playbook:** Create a Playbook where the condition is `Data Type = [Your New PHI SDD]`.
4.  **Action:** Set the action to **"Classify: PHI"** and **"Notify: Privacy Team"**.

**Operational Tip:** Use this SDD to drive your **Governance Dashboard**. By reporting on this specific SDD, you can show leadership exactly where your highest-risk, "true" PHI resides, separate from general PII.