Auto-Referral builds on Auto-Save by creating a structured referral record and routing it to the right provider or queue.
It’s ideal for clinics that handle high volumes of incoming referrals from external physicians or facilities.
How it works
When a document is identified as a referral:
- The document classification confidence meets your threshold.
- The system verifies or creates a patient record (if paired with Auto-Chart).
- Phelix extracts referring provider and procedure details, verifying them against your rules.
- If all parameters pass, a referral record is automatically created in the EMR and assigned per your routing setup.
📸 Screenshot Reference: Auto-Referral Details configuration modal
Key parameters
- Document Classification Confidence – The minimum confidence required for recognizing the document as a referral.
- Patient Match By – Determines which patient fields are used to confirm the match.
- Confidence on Patient Name and Unique Identifier – Separate controls for how strictly each field must align.
- Referring Provider Confidence – Determines how certain the system must be about who the referral is from.
- Location Match By – Required, optional, or not used.
- Procedure Confidence – How precisely procedure keywords must match before the referral is created.
Example
A dermatology clinic receives a new referral for “Jane Smith.”
The system recognizes it as a referral with 92% confidence, matches the patient using name and DOB, and identifies “Dr. Chen” as the referring provider.
The procedure “Skin Lesion Removal” meets your minimum match score.
Auto-Referral creates the referral record and routes it to the dermatology intake queue — no manual filing required.
Common reasons Auto-Referral holds back
- Patient record incomplete or unclear (e.g., missing DOB or unique ID)
- Low confidence on referring provider or procedure
- Conflicting information between referral letter and EMR
- Document classification confidence below threshold
When this happens, the fax remains in the inbox for review.
Adjusting your confidence levels or adding more training examples usually resolves repeated holdbacks.