eReferral Implementation Guide, published by Riziv-Inami Platform. This guide is not an authorized publication; it is the continuous build for version 2.0.0 built by the FHIR (HL7® FHIR® Standard) CI Build. This version is based on the current content of https://github.com/hl7-be/referral-uhmep/ and changes regularly. See the Directory of published versions

Complex Operations

This page describes business operations that cannot be represented by a single FHIR interaction. They must be implemented as a controlled sequence of API calls.

The individual operation pages remain the reference for each call payload. This page explains how to combine them.

Transfer An Assignment To Another Professional

Transferring an assignment from one professional to another is not a single API operation. The client removes the existing assignment and creates a new assignment.

The transfer is possible only while the old assignment can still be removed. If execution already started, the old task must be cancelled or completed according to the execution workflow instead of being deleted.

Sequence

Transfer An Internal Organization Assignment

When the assignment is internal to an organization, the new BePerformerTask must keep the same basedOn reference to the organization task.

{
  "resourceType": "Task",
  "meta": {
    "profile": [
      "https://www.ehealth.fgov.be/standards/fhir/referral/StructureDefinition/be-performer-task"
    ]
  },
  "status": "ready",
  "intent": "order",
  "focus": {
    "reference": "ServiceRequest/{serviceRequestId}"
  },
  "partOf": [
    {
      "reference": "Task/{referralTaskId}"
    }
  ],
  "basedOn": [
    {
      "reference": "Task/{organizationTaskId}"
    }
  ],
  "owner": {
    "reference": "PractitionerRole/{newSSIN}-{DISCIPLINE}"
  }
}

Operational Guidance

• Do not model multi-call business operations as a single client-side PATCH on multiple resources.
• After a multi-call workflow, read the task graph to verify the resulting state.
• If one step fails, use the returned OperationOutcome to decide whether the client can retry or must refresh the current state first.
• For operations that need atomicity across several resources, use a dedicated server operation when one exists, such as proposal approval or rejection.