The Unstructured Document retrieval format allows a provider to share information in standard formats which are generally not considered to have a machine readable structure, such as a PDF document containing a ‘Mental health crisis plan’.
Support for unstructured documents is intended to share existing document and record types, without the need for consumers and providers to develop and adopt new standards, resulting in reduced development and implementation time, quicker benefits for health care services and improved care for patients.
Pointer Retrieval Format Code and MIME Type
Format
The NRL pointer format code for this retrieval format is as follows:
| System | Code | Display |
|---|---|---|
| https://fhir.nhs.uk/STU3/CodeSystem/NRL-FormatCode-1 | urn:nhs-ic:unstructured | Unstructured Document |
Content-Type
The content-type of the unstructured document returned SHOULD be in the retrieval MIME type as described on the pointer metadata. Supported MIME types for pointer types which support unstructured documents are listed on the pointer types page.
Currently supported MIME types:
| Name | Type |
|---|---|
| application/pdf |
Retrieval Interaction
The Unstructured Document retrieval format MUST support the SSP Read retrieval interaction.
The endpoint URL included in the NRL pointer MUST require no additional/custom headers or parameters beyond those required by the SSP.
Retrieval Response
When successfully responding to an unstructured document request, the provider MUST return:
- an HTTP
200OK status code. - a payload conforming to the format and MIME included in the pointer reference. The unstructured document MUST NOT be wrapped in a structured data model such as a FHIR Binary resource.