Automating healthcare referral processing

The challenge

Patient referrals are the lifeblood of specialist clinics. But the way they arrive is stuck in the 1990s.

This outpatient clinic with 12 specialists received referrals from dozens of general practitioners — some by fax-to-email, others as scanned PDF attachments, a few through portal notifications, and occasionally just a free-text email with patient details buried in the body.

The intake coordinator spent half her day opening emails, downloading attachments, reading through referral letters, manually entering patient data into the scheduling system, and routing each case to the appropriate specialist.

Urgent referrals sometimes sat unnoticed in the queue for hours. Non-urgent ones occasionally got lost entirely. Patients waited longer, GPs got frustrated, and the clinic left appointment slots empty.

The solution

They set up specialty-specific intake aliases through EmailConnect. Each department gets its own address: cardiology@ingest.clinic.eu, ortho@ingest.clinic.eu, neurology@ingest.clinic.eu — plus a general referrals@ingest.clinic.eu for unspecified cases.

GPs send their referral letters to the relevant specialty address. EmailConnect extracts the email body, all attachments, sender details, and urgency indicators — then pushes everything as structured JSON to their patient management webhook.

Their automation layer does the heavy lifting: parsing patient demographics from attachments, flagging urgent cases for immediate triage, and pre-populating the scheduling system so the coordinator only needs to confirm and assign a time slot.

Step 1: EmailConnect delivers the parsed email

When a GP sends a referral to the specialty alias, EmailConnect parses everything and fires the webhook:

{
  "message": {
    "date": "2024-09-15T08:45:00.000Z",
    "sender": {
      "name": "Dr. A. de Vries",
      "email": "a.devries@huisarts-centrum.nl"
    },
    "recipient": {
      "email": "cardiology@ingest.clinic.eu"
    },
    "subject": "Referral — suspected atrial fibrillation",
    "content": {
      "text": "Dear colleagues,\n\nI'd like to refer my patient for cardiology evaluation. Irregular rhythm on ECG, patient reports palpitations for 3 weeks. Semi-urgent.\n\nKind regards,\nDr. A. de Vries\nHuisartsenpraktijk Centrum",
      "markdown": "Dear colleagues,\n\nI'd like to refer my patient for cardiology evaluation. Irregular rhythm on ECG, patient reports palpitations for 3 weeks. **Semi-urgent.**"
    },
    "attachments": [
      {
        "filename": "referral_letter_devries.pdf",
        "contentType": "application/pdf",
        "size": 838860,
        "url": "https://s3.eu-central-1.amazonaws.com/..."
      },
      {
        "filename": "ecg_results_20240915.pdf",
        "contentType": "application/pdf",
        "size": 1258291,
        "url": "https://s3.eu-central-1.amazonaws.com/..."
      }
    ]
  },
  "classification": {
    "type": "normal",
    "signals": ["has_attachment"],
    "confidence": "definite"
  }
}

Showing relevant fields only. See the full webhook payload reference →

Step 2: The triage workflow processes the referral

Their automation receives the webhook and applies triage logic:

• Department routing — the recipient alias (cardiology@) determines the specialty queue
• Urgency detection — "semi-urgent" in the body triggers a 48-hour scheduling target
• Attachment classification — referral letters separated from diagnostic results
• Scheduling pre-fill — GP details and referral reason pre-populated in the booking system
• Duplicate detection — sender + subject + date checked to prevent double processing

Step 3: The result

The patient management system is updated and everyone is notified:

{
  "referral": "REF-2024-3847",
  "department": "cardiology",
  "urgency": "semi-urgent — 48hr scheduling target",
  "referring_gp": "Dr. A. de Vries — Huisartsenpraktijk Centrum",
  "documents_stored": [
    { "file": "referral_letter_devries.pdf", "tagged": "referral_letter" },
    { "file": "ecg_results_20240915.pdf", "tagged": "diagnostic_result" }
  ],
  "actions_taken": [
    "Routed to cardiology department queue",
    "Patient record pre-populated in scheduling system",
    "Acknowledgment email sent to Dr. de Vries",
    "Cardiologist Dr. Bakker notified"
  ]
}

The results

Referral processing went from a half-day manual workflow to near-instant automated triage. Patient data could not leave the EU — a hard requirement. EmailConnect's EU-sovereign infrastructure ensures all referral processing stays within European borders, with no Cloud Act exposure, full GDPR compliance, and audit trails that satisfy healthcare data protection regulations.

• Urgent referrals triaged in minutes instead of sitting in a queue for hours
• GPs receive instant acknowledgment — improving inter-provider relationships
• No more lost referrals — every email tracked from receipt to appointment
• Scheduling pre-populated so coordinators confirm rather than enter data
• Full audit trail for healthcare compliance and quality reviews