Australian Digital Health Agency FHIR Implementation Guide 1.0.0

Australian Digital Health Agency FHIR Implementation Guide. This is the current published version. It is based on FHIR® version 4.0.1. See the Directory of published versions for a list of available versions.

Guidance

Business identifiers

“Business” identifiers are used extensively in ADHA profiles to consistently identify real world entities across systems, contexts of use, and other formats (e.g. HL7 v2 , CDA , XDS, and many more).

For ADHA profiles, the following identifier elements are populated with business identifiers:

  • Device.identifier
  • DiagnosticReport.identifier
  • HealthcareService.identifier
  • MedicationDispense.identifier
  • MedicationRequest.identifier
  • Organization.identifier
  • Patient.identifier
  • RelatedPerson.identifier
  • Practitioner.identifier
  • PractitionerRole.identifier
  • ServiceRequest.identifier

Business identifiers will typically be a national identifier (ABN, Medicare Provider, IHI), registry / exchange service identifier (ETP, eRx), or local identifier (MRN, Placer Identifier).

HL7 AU Australian Base Implementation Guide publishes and maintains rules on how to exchange various business identifiers in Australia as a set of Identifier data type profiles, e.g. AU PBS Prescriber Number. An identifier data element SHALL conform to the applicable HL7 AU Identifier profile.

While national and registry / exchange service identifiers will define the namespace to use when sending an identifier, a local identifier requires the organisation to define their own namespace when exchanging identifiers they manage.

When constructing a local identifier it is preferable that an organisation uses their own local system identifier namespace but if that is not available then an organisation can use their HPI-O or ABN to construct a legal, globally unique identifier system for their local identifiers.

HPI-O scoped identifiers

HPI-O scoped identifiers enable exchange of an organisation’s local identifiers for items such as a patient medical record or a pathology report by combining a dedicated Agency published namespace and their HPI-O to construct a legal, globally unique identifier system for their local identifiers.

The full list of available identifier namespaces can be found by browsing the ns.electronichealth.net.au identifier namespaces; there are several HPI-O scoped identifier namespaces available:

  • http://ns.electronichealth.net.au/id/hpio-scoped/accessionnumber/1.0
  • http://ns.electronichealth.net.au/id/hpio-scoped/dispense/1.0
  • http://ns.electronichealth.net.au/id/hpio-scoped/medicalrecord/1.0
  • http://ns.electronichealth.net.au/id/hpio-scoped/order/1.0
  • http://ns.electronichealth.net.au/id/hpio-scoped/prescription/1.0
  • http://ns.electronichealth.net.au/id/hpio-scoped/report/1.0
  • http://ns.electronichealth.net.au/id/hpio-scoped/service-provider-individual/1.0

There are four parts to using a HPI-O scoped identifier in FHIR: system, value, assigner and depending on the identifier profile requirements, a coded type.

The system value is constructed in the format of [baseURL]/HPI-O, e.g.:

"system" : "http://ns.electronichealth.net.au/id/hpio-scoped/service-provider-individual/1.0/8003621566699776"

The value contains the local identifier, e.g.:

"value" : "AMC-GA-001"

The assigner contains the name of the organisation that issues or manages the identifier, e.g.:

assigner" : {
    "display" : "Algregster Medical Center QLD"
}

Example: PractitionerRole resource with an employee number (local identifier)

{
  "resourceType" : "PractitionerRole",
    ...
    "identifier" : [
      {
        "type" : {
          "coding" : [
            {
              "system" : "http://terminology.hl7.org/CodeSystem/v2-0203",
              "code" : "EI",
              "display" : "Employee number"
              }
            ],
          "text" : "Employee Number"
      },
      "system" : "http://ns.electronichealth.net.au/id/hpio-scoped/service-provider-individual/1.0/8003621566699776",
      "value" : "AMC-GA-001",
      "assigner" : {
        "display" : "Algregster Medical Center QLD"
    },
  ...
}

ABN scoped identifier

ABN scoped identifiers enable exchange of an organisation’s local identifiers for items such as a patient medical record by combining a dedicated Agency published namespace and their ABN to construct a legal, globally unique identifier system for their local identifiers.

The full list of available identifier namespaces can be found by browsing the ns.electronichealth.net.au identifier namespaces; there are two ABN-scoped identifier namespaces available:

  • http://ns.electronichealth.net.au/id/abn-scoped/medicalrecord/1.0
  • http://ns.electronichealth.net.au/id/abn-scoped/service-provider-individual/1.0

There are four parts to using an ABN scoped identifier in FHIR: system, value, assigner and depending on the identifier profile requirements, a coded type.

The system value is constructed in the format of [baseURL]/ABN, e.g.:

"system" : "http://ns.electronichealth.net.au/id/abn-scoped/medicalrecord/1.0/004085616"

The value contains the local identifier, e.g.:

"value" : "123456"

The assigner contains the name of the organisation that issues or manages the identifier, e.g.:

assigner" : {
    "display" : "Test Hospital Org 1"
}

Example: Patient resource with a medical record number (local identifier)

{
  "resourceType" : "Patient",
    ...
      "identifier" : [
        {
        "type" : {
          "coding" : [
            {
              "system" : "http://terminology.hl7.org/CodeSystem/v2-0203",
                     "code" : "MR",
                     "display" : "Medical record number"
                   }
                 ],
          "text" : "Medical Record Number"
        },
        "system" : "http://ns.electronichealth.net.au/id/abn-scoped/medicalrecord/1.0/004085616",
        "value" : "123456",
        "assigner" : {
          "display" : "Test Hospital Org 1"
    },
  ...
}

Addresses

All Australian addresses SHOULD conform to the Australian Address profile.

References between resources

References between resources in ADHA profiles are supported as reference (literal reference), identifier (logical reference), and display (text description of target).

ADHA profiles may include constraints on elements of Reference type that limit what is considered valid. For example, the profile ADHA Record of Claim against MBS or DVA limits what is considered valid for the element ExplanationOfBenefit.referral by mandating ExplanationOfBenefit.referral.reference to enforce population of a literal reference if the referral element is provided.

If an identifier (logical reference) is supplied, that identifier SHALL be populated with a meaningful business identifier according to the section on Business identifiers that identifies the logical entity across systems, contexts of use, and other formats (e.g. HL7 v2 , CDA , XDS, and many more).

References to a patient

References to a patient SHALL populate identifier and when applicable populate a reference to a Patient resource:

  • identifier SHOULD be populated with a verified IHI
  • if referencing a specific Patient resource instance reference SHALL be populated and it SHALL resolve

Example: Observation resource with a Reference to a Patient resource as identifier and reference

{
  "resourceType" : "Observation",
  ...
  "subject" : {
    "reference" : "Patient/88923001",
    "identifier" : {
      "type" : {
        "coding" : [
          {
            "system" : "http://terminology.hl7.org/CodeSystem/v2-0203",
            "code" : "NI",
            "display" : "National unique individual identifier"
          }
        ],
        "text" : "IHI"
      },
      "system" : "http://ns.electronichealth.net.au/id/hi/ihi/1.0",
      "value" : "8003608000228445"
    },
  ...
}

Contained resources

In some circumstances, the content referred to in the resource reference does not have an independent existence apart from the resource that contains it - it cannot be identified independently, and nor can it have its own independent transaction scope. For example, use of a Medication resource to represent medicinal product identification within the context of a MedicationRequest. In these circumstances the resource should be contained.

If referencing a contained resource, both the contained resource and the referencing resource SHALL conform to an ADHA profile. Further guidance about the general use case for contained resources can be found in the base FHIR specification.

In ADHA profiles:

  • An ADHA PBS Claim Item (MedicationRequest resource) SHOULD be contained within the ADHA Record of Claim against PBS or RPBS (ExplanationOfBenefit resource).
  • An ADHA Organ or Tissue for Donation BodyStructure (BodyStructure resource) SHOULD be contained within the ADHA Record of Consent from Australian Organ Donor Register (Consent resource).
  • Systems constructing a resource that represents medication or body structure information are encouraged to make use of contained resources.
    • Operations on Medication resources are expected to be within the context of a referencing resource query such as an ExplanationOfBenefit, Flag, MedicationAdministration, MedicationDispense, MedicationRequest or MedicationStatement.
    • Operations on BodyStructure resources are expected to be within the context of a referencing resource query such as a Consent, DiagnosticReport, Observation or ServiceRequest.
  • Otherwise, when responding to a query, servers should not use inline contained resources to represent the returned data.

Missing data

There are situations when information for a particular data element is missing and the source system does not know the reason for the absence of data. If the source system does not have data for an element with a minimum cardinality = 0 (including elements labelled MustSupport), the data element SHALL be omitted from the resource. If the data element is a Mandatory element (in other words, where the minimum cardinality is > 0), it SHALL be present even if the source system does not have data. The core specification provides guidance for what to do in this situation, which is summarised below:

  1. For non-coded data elements including type Reference,
    • use the DataAbsentReason extension in the data type if the ADHA profile for that resource does not require a child element.
    • if the ADHA profile mandates a child element, such as a valid identifier or reference, then the resource must contain that element otherwise the instance will not be conformant.
    • use the code unknown where the value is expected to exist but is not known.

    Example: ExplanationOfBenefit resource where the patient’s insurance coverage is not available

    {
  "resourceType" : "ExplanationOfBenefit",
       ...
       "outcome" : "complete",
         "insurance" : [
           {
             "focal" : true,
             "coverage" : {
               "extension" : [
                 {
                   "url" : "http://hl7.org/fhir/StructureDefinition/data-absent-reason",
                   "valueCode" : "unknown"
                 }
               ]
             }
           }
         ],
         ...
     }
  2. For coded data elements:
    • example, preferred, or extensible binding strengths (CodeableConcept or Coding datatypes):
      • if the source system has text but no coded data, only the text element is used.
        • for Coding datatypes, the text only data is represented as a display element.
      • if there is neither text or coded data:
        • the appropriate “unknown” concept code SHALL be present if the binding strength is extensible.
        • if the value set does not have an appropriate “unknown” concept code, use unknown from the DataAbsentReason Code System.

        Example: AllergyIntolerance resource where the manifestation is unknown

        ...
"reaction" : [
  {
    "manifestation" : [
      {
        "coding" : [
          {
            "system" : "http://terminology.hl7.org/CodeSystem/data-absent-reason",
            "code" : "unknown",
            "display" : "unknown"
          }
        ]
      }
    ]
  }
]
...
    • required binding strength (CodeableConcept or code datatypes):
      • the appropriate “unknown” concept code SHALL be present if available.

      • if the value set does not have the appropriate “unknown” concept code you must use a concept from the value set otherwise the instance will not be conformant.

        • For ADHA Core profiles, the following mandatory or conditionally mandatory* status elements with required binding have no appropriate “unknown” concept code:
          • AllergyIntolerance.clinicalStatus*
          • Condition.clinicalStatus*
          • Composition.status
          • DiagnosticReport.status
          • DocumentReference.status
          • ExplanationOfBenefit.status
          • Immunization.status
          • List.status

        *The clinicalStatus element is conditionally mandatory based on resource specific constraints.

Suppressed data

In some circumstances, specific pieces of data may be hidden due to security or privacy reasons. Elements with a minimum cardinality = 0 (including elements labelled MustSupport) SHALL be omitted from the resource if they are suppressed.

For mandatory elements (minimum cardinality is > 0), the element SHALL be populated but it may exceed the data receiver’s access rights to know that the data is suppressed:

Extensibility – “additional” elements

A sending system may send “additional” elements beyond those labelled with MustSupport in an ADHA profile. Additional elements allow local requirements to be reflected including technical and workflow context for the resource, and extending the health information supported in exchanges. For this reason extensibility is generally allowed in ADHA profiles, and only in some use case profiles are the rules tightened to limit the nature of additional information that can be sent.

Depending on local requirements, a receiving or persisting system may ignore these “additional” elements, may treat the data as for rendering only, or be capable of recognising and using the element. A resource exchanged containing an additional element SHALL conform to ADHA conformance requirements. Requirements applicable to handling “additional” elements are summarised below:

  • an “additional” element SHALL be part of a supported resource, it may be a canonical element or an extension.
  • an “additional” element SHALL conform to the HL7 FHIR standard.
  • a resource referenced by an “additional” element SHALL conform to an ADHA profile and the HL7 FHIR standard.
  • where the “additional” element is an extension

Orphaned resources, i.e. not referenced by an element in supported resource, SHALL NOT be allowed.

System obligations on handling additional elements are:

  • systems that construct or send information SHALL ensure that a resource meets all applicable ADHA conformance requirements
  • systems that receive or persist information, when sent a resource with an “additional” set of elements:
    • SHALL meaningfully process the MustSupport elements; depending on local requirements this may mean display, persist, index, or action in an event or request workflow.
    • SHOULD persist additional elements where the system is capable of doing so, and the additional element is not a modifier extension.
    • MAY choose to ignore the additional elements.
    • MAY choose to reject non-conformant resources but are not required to.

Medicine information

The FHIR standard defines the following resources for exchanging medicine information:

ADHA Core Medication is profiled to support medicinal product identification in an Australian healthcare context. ADHA profiles of MedicationStatement (with ADHA Core Medication) are used to support summary statements of medicine use. ADHA profiles of MedicationAdministration (with ADHA Core Medication) are used to support medication chart and other administration use cases. ADHA profiles of MedicationDispense (with ADHA Core Medication) are used to support to support dispense records and ePrescribing use cases. ADHA profiles of MedicationRequest (with ADHA Core Medication) are used to support prescription, ordering, and ePrescribing use cases.

Medicinal product identification

For extemporaneous medications, the medication code is the mandatory primary mechanism to identify a medicine but contain only a text list of ingredients or it may be a code from a medicines terminology.

For non-extemporaneous medications, the medication code (or set of codes) is the mandatory primary mechanism to identify a medicine and its defining attributes (by terminology lookup) including form and strength.

Australian Medicines Terminology (AMT) is the national terminology for identification and naming of medicines in clinical systems for Australia. The AMT is published monthly to include new items on the Australian Register of Therapeutic Goods from the TGA, as well as items listed on the Pharmaceutical Benefits Scheme. The AMT is published as part of SNOMED CT-AU (Australian edition of SNOMED CT) and can be downloaded in a variety of formats from the National Clinical Terminology Service (NCTS).

In addition to the medication code, the majority of use cases support exchange of structured medicine information as separate data elements covering brand name, generic name, item form and strength, and manufacturer.

These data elements may be supported as coded, or text, and systems are likely to use a combination of coded and text elements when constructing a Medication resource. The guidance for how to support coded or text is summarised below:

  1. For coded support for brand name, generic name, manufacturer, item form and strength:

    Example: Medication with coded brand name, generic name, manufacturer, item form and strength

     {
   "resourceType": "Medication",
     ...
     "code": {
       "coding": [
         {
           "extension": [
             {
               "url": "http://hl7.org.au/fhir/StructureDefinition/medication-type",
               "valueCoding": {
                 "system": "http://terminology.hl7.org.au/CodeSystem/medication-type",
                 "code": "UPD",
                 "display": "Unbranded product with no strengths or form"
               }
             }
           ],
           "system": "http://pbs.gov.au/code/item",
           "code": "02647H",
           "display": "BENZYLPENICILLIN"
         },
         {
           "extension": [
             {
               "url": "http://hl7.org.au/fhir/StructureDefinition/medication-type",
               "valueCoding": {
                 "system": "http://terminology.hl7.org.au/CodeSystem/medication-type",
                 "code": "BPD",
                 "display": "Branded product with no strengths or form"
               }
             }
           ],
           "system": "http://snomed.info/sct",
           "code": "3539011000036105",
           "display": "Benpen"
         },
         {
           "extension": [
             {
               "url": "http://hl7.org.au/fhir/StructureDefinition/medication-type",
               "valueCoding": {
                 "system": "http://terminology.hl7.org.au/CodeSystem/medication-type",
                 "code": "UPDSF",
                 "display": "Unbranded product with strengths and form"
               }
             }
           ],
           "system": "http://snomed.info/sct",
           "code": "32753011000036104",
           "display": "benzylpenicillin 3 g injection, 1 vial"
         },
         {
           "extension": [
             {
               "url": "http://hl7.org.au/fhir/StructureDefinition/medication-type",
               "valueCoding": {
                 "system": "http://terminology.hl7.org.au/CodeSystem/medication-type",
                 "code": "BPDSF",
                 "display": "Branded product with strengths and form"
               }
             }
           ],
           "system": "http://snomed.info/sct",
           "code": "32328011000036106",
           "display": "Benpen 3 g powder for injection, 1 vial"
         }
       ]
     },
     "manufacturer": {
       "identifier": {
         "system": "http://pbs.gov.au/code/manufacturer",
         "value": "CS"
       }
     },
     "form": {
       "coding": [
         {
           "system": "http://snomed.info/sct",
           "code": "129011000036109",
           "display": "injection"
         }
       ],
       "text": "Injection"
     },
     "ingredient": [
       {
         "itemCodeableConcept": {
           "coding": [
             {
               "system": "http://snomed.info/sct",
               "code": "1849011000036104",
               "display": "benzylpenicillin"
             }
           ]
         },
         "strength": {
           "numerator": {
             "value": 3,
             "unit": "g"
           },
           "denominator": {
             "value": 1,
             "unit": "unit"
           }
         }
       }
     ]
 }
  2. For non-coded support for brand name, generic name, manufacturer, item form and strength:

    Example: Medication with text only brand name, generic name, item form and strength

    {
  "resourceType": "Medication",
   ...
   "extension": [
     {
       "url": "http://hl7.org.au/fhir/StructureDefinition/medication-generic-name",
       "valueString": "Benzylpenicillin"
     },
     {
       "url": "http://hl7.org.au/fhir/StructureDefinition/medication-brand-name",
       "valueString": "Benpen"
     }
   ],
   "code": {
     "text": "Benpen 3 g powder for injection, 1 vial"
   },
   "manufacturer": {
     "display": "Seqirus"
   }
}

Representing communication preferences

Patient

The table below provides guidance on representing communication preferences for a patient. Blank cells in the table indicate that the given element is absent from the resource.

Scenario communication.language communication.preferred extension:interpreterRequired Notes
Preferred language is English No element sent, as per the guidance in the Comments of Patient.communication
Preferred language is other than English language.coding 'true'
Interpreter required, language is known language.coding 'true' 'true'
Interpreter required, language is not known 'true'
Interpreter is not required 'false'
Communicates with multiple languages language.coding Each language instantiated in separate communication nodes; communication.preferred and extension:interpreterRequired may be sent as needed.

Example: Patient resource with interpreter required and language is known

{
  "resourceType" : "Patient",
    ...
      "extension" : [
        {
          "url" : "http://hl7.org/fhir/StructureDefinition/patient-interpreterRequired",
          "valueBoolean" : true
        }          
       ]
    },
    ...
    "communication" : [
    {
      "language" : {
        "coding" : [
          {
            "system" : "urn:ietf:bcp:47",
            "code" : "yue"
          }
        ],
        "text" : "Cantonese"
      },
      "preferred" : true
    }
  ]
}

RelatedPerson

The table below provides guidance on representing communication preferences for a related person. Blank cells in the table indicate that the given element is absent from the resource.

Scenario communication.language communication.preferred Notes
Preferred language is English No element sent, as per the guidance in the Comments of RelatedPerson.communication
Preferred language is other than English language.coding 'true'
Communicates with multiple languages language.coding Each language instantiated in separate communication nodes; communication.preferred may be sent as needed.