Coordinated Registry Network (CRN) HL7 Implementation Guide (IG) Workgroup (WG)

Meeting Date & Time

Tuesday, May 21, 2019 at 2:00 pm ET

Attendees

Attendees

Present/Absent

Attendees

Present/Absent

Becky Angeles

P

Behnaz  Minaei

X

Nagesh Bashyam (Dragon)

P

JaWanna Henry

X

Gayathri Jayawardena

P

Myron Finseth

P

Richard Ballew

P

Vaishnavi Rao

P

Rob Samples

X

Ioana Singureanu

X

Terrie Reed

X

Marti Velezis

P

Lisa Lang

X

Patrick McLaughlin

P

Abdullah Rafiqi

P

Robin Taylor

P

Christina Nguyen

X

Lindsey Hoggle

X

Discussion                                                                    

HL7 WGM Activities Summary

  • TCRN STU Ballot Results
    • 119 votes
      • Affirmative 11
      • Negative 11
      • Abstain 84
      • No votes 13
    • 40 comments (includes FHIR tracker comments)

 

CRN Implementation Guide (IG) Comment Reconciliation

  • FHIR Tracker
    • Query for CRN specification, May 2019 ballot to view comments
    • Typos are auto-approved
  • Comment Review
    • 21319 - Submitted by Marti Velezis . Suggested remove and replace text in the additional Profile specific implementation guidance with the following, “The deviceDefinition-crn profile is intended to be used for reporting the use or implant of a medical device when the following information is available in the source system”.
      • Marti suggested the IG needs to address device information provided by both provider and patient.  If the UDI is present, the DI would be parsed and used to collect the device definition.  If it came from a UDI, then all the info should come from the device profile and if it came from a patient, you won’t have the UDI and should therefore use device definition.  Marti suggested additional device information should be pulled from the GUDID database, if available, and populated in the device definition.  A consistent way is needed to convey the information moving forward.  If UDI is always required, then what we put in IG makes sense; however if a patient reports historical data, we need a way to address that scenario as well.  Manufacturer, device name, model number, and type should be populated from GUDID.  If those are also in the device resource, then only the device resource should be used to collect that info. 
      • Dragon asked if the Device Definition should be removed.  Marti thought it should remain for patient-reported devices, as this historical data might not be available in the EHR.  Dragon asked how the device information is attained of the UDI is unavailable.  Marti explained that this could be reported by patient.  For example the patient reports having a pacemaker.  This is sufficient enough from a historical standpoint to record.
      • Richard Ballew noted UDI information was to be captured at that time; however the understanding was that FDA had this aspect identified.  He suggested something is in the works within HL7 community to address the UDI capture.
      • All implantable devices should have a UDI; however in the case of historical device information (e.g., pacemaker, hip replacement, etc.), the UDI may be unknown.  This requirement needs to be addressed. 
      • Next Steps: Table this comment for now.  Wait for ballot reconciliation of UDI pattern ballot (June 3 rd ).
    • 21315 - All references to the GUDID database need to be clarified to AccessGUDID database managed by the National Library of Medicine at: https://accessgudid.nlm.nih.gov/
      • Resolution: Change required referencing AccessGUDID instead of GUDID and sending for block vote.
    • 20958 - Add section numbering in IG. 
      • Resolution: Change required to correct and send for block vote.
    • 20966 - Change “NLM’s CDE Repository” to “NIH CDE Repository”.
      • Resolution: Change required to correct and send for block vote.
    • 20967 - The Questionnaire Resource is a demonstration of what the elements might look like in a questionnaire, not a usable questionnaire.  Dragon asked if the reference should be removed or revised to indicate, “For visualization of the CDEs refer to this resource”. 
      • Robin Taylor will consider offline and report back to the group.
    • 21013 - Patient communication language - The template refers to both communication language and preferred language.  
      • Resolution: For pat_pref_language_spoken add mapping to Patient.communication.preferred to indicate that the language represented in Patient.communication.language is a preferred language.
    • 21014 - Suggestion to include full set of Common Data Elements in the IG rather than a separate spreadsheet in Google drive. 
      • Resolution: Add the complete data element spreadsheet to the ‘Overview’ and the ‘Mappings and Profiles’ pages.
    • 21015 - Suggestion to include LOINC and other codes whenever they have been specified as a Question Code in the template. 
      • Create a value set and include the value set in the IG or host externally and reference the external links.  Dragon noted the value sets cannot be updated as regularly if included in the IG.
      • Richard suggested with the value sets, we can reference grouping OIDs and update them independently.  This is accessible through VSAC.  A UMLS license is needed to access.
      • Resolution: Link to the VSAC FHIR API for value sets in the Terminology section.  Include a mention of the license requirements to access the value sets.
    • 21016 - If AdverseEvent is “not specific enough currently” should it be removed from the current IG?
      • Richard explained that the Clinical Working Groups noted that determining relatedness of device-related adverse event (AE) is a judgement call that is viewed as burdensome if put on provider.  This was focused on collecting and using available AE capture information, recognizing that it was likely to mature as it went through this round. 
      • Richard suggested discussions were focused on not reinventing AE reporting, but rather the expectation was that the AE FHIR Resource would be used.
      • Dragon noted this AE FHIR Resource is comprised of a large group of data elements and specific guidance about which elements to capture and which value sets to use, is needed to narrow this list.
      • Resolution: Richard suggested this issue should be deferred to Clinical Working Groups.  He will follow up with Project Lead offline to address this issue.

 

 

Next Steps

  • The Support Team will invite all the commenters to the next week’s call to continue the discussion.