Skip to end of metadata
Go to start of metadata

Scott Gordon - FDA

To the Co-Chairs of the HL7 Biomedical Research and Regulation (BR&R) Workgroup,

In response to a request by the BR&R workgroup for statements on the impact and importance of the release of HL7 FHIR R5 to FHIR initiatives, I submit this comment as the Program Manager for FDA’s efforts to standardize Pharmaceutical Quality/Chemistry, Manufacturing & Controls data for submission to the Food and Drug Administration (US FDA):

US FDA is in the process of implementing the exchange of data elements and terminologies for information commonly used and submitted in support of drug product applications in eCTD Module 3.  This is the Pharmaceutical Quality/Chemistry, Manufacturing & Controls (PQ/CMC) Project in the HL7 Biomedical Research & Regulated (BR&R) work group.  The PSS is here https://confluence.hl7.org/display/BRR/Pharmaceutical+Quality+%28PQ%29+PSS

FHIR was selected for that exchange mechanism after HL7 V3 SPL proved to be inadequate and HL7 indicated that initiating new standards in V3 was not supported.  A preliminary schedule for issuing regulatory guidance was planned based on published statements from HL7, “R5 would be expected to be published in Q3 2020” (https://onfhir.hl7.org/2019/01/20/fhir-r5-roadmap/).

FDA’s FHIR design and implementation work is entirely dependent on R5.  The resources that PQ/CMC project requires are only available in R5.  They include: MedicinalProductDefinition, Ingredient, ManufacturedItemDefinition, PackagedProductDefinition, AdministrableProductDefinition and SubstanceDefinition, to name a few.

Delaying FHIR R5 (or at minimum an STU update making all the needed resources available) beyond Q3 2020 has a direct impact on FDA PQ/CMC planned timeline.  The project is tasked with providing a timely implementation of the provisions from the 2012 Food and Drug Administration Safety and Innovation Act (FDASIA) (Pub. L. 112-144), which authorized the Agency to require certain submissions to be in a specified electronic format. 

The PQ/CMC project urges the FHIR management to provide timely access to FHIR R5 to minimize impact to FDA regulatory plans of FHIR based submission of structured PQ/CMC data from the industry to the FDA.

Thank you for the opportunity to raise our concerns and recommendations.

Peter Bomberg - Health Canada

HC is aiming to use FHIR messages as part of the replacement of the Electronic Licencing & Inspection System (ELIS) to support the Regulatory Operations and Enforcement Branch (ROEB), however as per our internal guidance we cannot ask industry to leverage the Work In Progress (i.e the FHIR build) specification, therefore we require a stable release i.e. R5.

While several resources are required, they can be characterized RegulatedAuthorization, Procedure, List, DocumentReference and Organization and all resources referenced by these, this includes items such as Practitioner, PractitionerRole, Group, etc.

Rik Smithies EMA and UK dm+d

EMA have a strong need to get R5 published.

The systems they are building now are based on the R5 Medication Definition resources.

Stakeholders want to see a formal release with these resources in it.

All the resources have changed names since R4 (which were mostly requested by FMG).

So currently they don’t really exist in R4 -  they are there with old names but that doesn’t look very good when convincing people these are real.

The R5 resources have lots of changes since R4. Lots happened before they went to FMM1 (which itself is an important milestone for stakeholders - they were FMM 0 in R4, with the old names)

And there have been quite a few more changes since, via all these Jiras you are helping get approved. The changes are needed for EMA or FDAs data.

Another reason is that the uk dm+d has been mapped to R5, and some changes made to accommodate it, so it’s much better fit than the R4 set.

In a nutshell the Medication definition resources only really exist properly in R5 and at least 3 projects need them (EMA, FDA, and some work I am doing on dm+d over here - not directly for NHS, but anyway).

  • No labels