Question about FHIR lab modelling

Also, consult the v2 - FHIR project, their confluence page can be found at: confluence.hl7.org/display/OO/2-To-FHIR+Project

Their conference calls are at 11 am ET on Thursdays, but there will not be a call next week due to the virtual HL7 conference.

Lorraine Constable

Makes sense. The concept of "collection of ServiceRequests" is similar to ORC to support a group of requests in one requisition, but the challenge is how to support OBRs in the ORC. If we map OBR to DiagnosticReport, we can't have DiagnosticReport reference other DiagnosticReports. If we map to ServiceRequest, ServiceRequest.status does not support common OBR status values (partial, final, corrected). Should we look at mapping OBRs to observations?

What constitutes a distinct DiagnosticReport is sort of up to the business process. It may be "one DiagnosticReport per ServiceRequest", "one DiagnosticReport per ServiceRequest.requisition - i.e. a collection of ServiceRequests", or it might be split other ways based on lab processes or regulations. Whether you transmit a bunch of DiagnosticReports together in a Bundle (e.g. transaction or document) or send them separately is also a business choice, though here there'd ideally be some standardization driven by local implementation guides.

Hi Fang,

Hi Fang,

The order catalog work provides a framework for publishing a lab compendium of tests that you can order, and how to describe the definition of what can be ordered. An actual order should be a Service Request, and the report is a Diagnostic Report that contains Observations. More detail can be obtained from the HL7 Orders and Observations (OO) workgroup, either on the listserv (Cette adresse courriel est protégée contre les robots spammeurs. Vous devez activer le JavaScript pour la visualiser.), which you need to join in order to send messages, or on Zulip at chat.fhir.org

Lorraine Constable
OO Co-chair
Cette adresse courriel est protégée contre les robots spammeurs. Vous devez activer le JavaScript pour la visualiser.
780 951 4853
Mountain Standard Time (UTC -7)

Hi Fang, you may explore an Order Catalog IG to see how others built it:
build.fhir.org/ig/HL7/fhir-order-catalog/index.html

Hello. Does the working group have any recommendation on the best practice of FHIR modelling for returning lab reports? Typically a lab report consists of test request and each test request consists of lab results. Should a lab report be mapped to one DiagnosticReport or a bundle of DiagnosticReports? Should a test request (i.e. OBR) be mapped to a DiagnosticReport or ServiceRequest? I can see implementers doing it both ways and causing inconsistencies. If the baseline can provide guidance on the best approach, that would be really helpful.

Thank you!

Fang Cao