Augmented unique device identifier (AUDI)

Mission & Background


The mission of the Augmented UDI (AUDI) workgroup is to provide the framework for best practices in expanding the UDI-associated device system to manage clinically significant attributes not currently found in the GUDID.



Unique device identifiers for coronary stent postmarket surveillance and research: a report from the Food and Drug Administration Medical Device Epidemiology Network Unique Device Identifier demonstration.

1. Rename the concept of the SUDID

  • “Supplemental” suggests lack of importance – suggestions include AUDI db (Augmented UDI database), EDAD (Extended Device Attributes Database) HVDAD (High Value Device Attributes Database) or CRD3 (Clinical Relevant Device Data Database – but that would be “crude” versus “good” – as in GUDID)
  • Starting point: clinically high value attributes / descriptors of devices that are not captured (or not captured systematically) as discrete data in the GUDID (also see below re: other use cases)

2. Perspectives of data users

  • Clinicians identified the need for clinically relevant, device specific attributes
  • Mercy Demonstration used a multi-stakeholder group including industry and clinicians to identify clinically necessary attributes (not just a regulatory perspective)
  • Other stakeholders may need representation
  • AUDI needs flexibility to add or delete attributes or change their definitions

3. Perspectives of FDA and NLM

  • What is the FDA / NLM requesting, and what would be useful to return to FDA / NLM?
  • What can FDA / NLM actually act upon, and what would nominally be out of bounds?
  • If out of bounds, what would be the process to bring recommendations in bounds?
  • NLM GUDID data model, potential / capacity to actually augment the GUDID database

4. Review (demonstration) of GUDID system

  • Standard interface at
  • Using the GUDID filters, what the filters are based on
  • JSON query via API, and what the API can / cannot do

5. Review the SUDID work of the Mercy demonstration project

  • Overall demonstration project: Drozda JP Jr. et al., Healthc (Amst). 2016 Jun;4(2):116-9.
  • SUDID component of the demonstration project: Tcheng JE et al., Am Heart J 2014;168:405-413.e2
    • 18 use cases (clinical care, supply chain management, consumer information, research, regulatory, and surveillance domains)
    • 9 SUDID coronary stent attributes (length, diameter, nonconventional property, structural material, coating(s), drug(s), strut thickness, expansion method, MRI compatibility)

6. Identify stakeholders who will be using the data, needs of those stakeholders

  • Build on the experience of the Mercy demonstration project (i.e., key clinically relevant data is not in the GUDID, or not in the GUDID systematically – e.g., stent attributes)
  • Consider additional key use cases
  • Identify principles for inclusion of a data element in the AUDI dataset (“what constitutes a relevant data element to include in the AUDI and who gets to determine it?”)

7. Review the Clinically Relevant Size data elements in the GUDID

  • See RAPID data elements spreadsheet, device data tab
  • See Appendix 1 for examples of coronary stent data returned from the GUDID
  • See Appendix 2 for description of observations about Clinically Relevant Size

8. Review relevance, need for device classification as a critical underpinning

  • Assumption: augmented data should be managed by class of device, not by individual device – is this correct?
  • Review classification schemas from GMDN, FDA, and SNOMED

9. Perspectives of industry

  • Supplying / updating / revising information in the GUDID will require time, effort, and resources. What does best state look like from an industry perspective in terms of providing / updating AUDI?
  • How is accuracy best assured, and who bears liability for mistakes?


10. Recommendations and artifacts

  • List of use cases to be evaluated when considering a data element for the AUDI (are the 18 use cases in the AHJ paper comprehensive?)
  • List of stakeholders to include in device by class AUDI workgroups
  • Recommendation for approach to device classification (i.e., derived GMDN, FDA, or SNOMED)
  • Recommendations for operational approach to identification and prioritization of device classes (i.e., overall organization and governance)
    • Responsibility for continuing activity – NEST, MDEpiNet, other?
    • FDA / NLM administrative resources to organize and manage
    • Funding of administration of program
  • Recommendations for operational approach to identifying key device attributes within a device class (i.e., governance of device class workgroups)
    • Principles about “what” and “who”, then adding the “how”
    • “Form” (or file format), process to submit attributes list to AUDI owner to be built
    • Notification of manufacturers to update data in GUDID
  • Technical recommendations
    • Hosting of the AUDI database - extension of NLM GUDID system or alternative?
    • Table structure for handling attributes, table relationships
  • AUDI process and procedure
    • Coordinating with FDA for instructions to manufacturers
    • Extending the GUDID API to return AUDI data
    • Adding device classification filters to GUDID UI, adding device classification query
    • Publication of device classification schema

11. Review and dissemination

  • Submission of recommendations to FDA and NLM
  • Posting to LUC, BUILD (Natalia Wilson), MDEpiNet, etc.
  • Posting via FDA channels
  • Manuscript opportunity

Teams and Member

AUDI Working Group members were asked to volunteer to lead or participate in one or more Teams; leaders from each Team will meet regularly with the leadership team to coordinate activities and recommendations across teams. Team leads are encouraged to interact with each other if/when they identify areas in which activities or recommendations need to be aligned or interfaces need to be considered.

Each Team will identify and articulate principles & proposed organizational process and structural components of plans to achieve the goals of their AUDI topic areas. The actual implementation of these plans is beyond the scope of the AUDI WG.  The expectation is that the WG leaders and members are ambassadors that will be champions of broad adoption and implementation of the AUDI recommendations.

The recommendations and plans from all Teams will be synthesized into a white paper, manuscript, slides, quick reference guides and other ‘marketing’ tools to facilitate dissemination and adoption.  Our stakeholder groups are the intended audience for these materials, including but not limited to government and regulatory agencies, Learning UDI Community, clinicians, analysts and researchers; device manufacturers, HIT vendors, medical device professional societies and many others.

Recommendations and plans will be shared with MDEpiNet leadership and stakeholders for endorsement.  They will then be shared with the MDEpiNet Pilot Project leaders and others who can exercise the recommendations on real projects; governing and hosting data platforms in which AUDI data elements for devices of interest to the pilot projects can be developed, managed and exchanged.

Scope & Strategy Team

Lead by Madris Tomes, Device Events

First Name Last Name Organization
H. Vernon Anderson Univ. of Texas Health Science Center Houston
Dennis Black Beckton-Dickinson
Denea Boyle Novant Health
Joe Drozda Mercy Health System
Hitinder Gurm University of Michigan
Robert Perry DoD
James Philips FMOLHS
CJ Rieser University of Virginia, MITRE
David Slotwiner Weill Cornell Medical College - New York Presbyterian Hospital


Governance & Funding Team

Lead by Brian Fortier, Aorta Medical

First Name Last Name Organization
Jodi Akin Hawthorne Effect, Inc.
Susan Broyles FDA Patient Representative
Hitinder Gurm University of Michigan
Melanie Hadlock Bard Peripheral Vascular
Roxana Mehran Mount Sinai Hospital
Zach Rothstein AdvaMed
Patricia Shrader Medtronic plc
Kweli Thompson Medtronic
Erin Williams CAMH/MITRE
Diana Zuckerman National Center for Health Research


Hosting & Structure Team

Lead by Patrick Lupinetti, First Databank

First Name Last Name Organization
John Gibson Mitre
Ted Heise Cook Medical
Stan Huff Intermountain
Graham Nichol University of Washington
Ahmed Saad MedStreaming LLC
Franco Sagliocca Mount Sinai Health System
Michael Simanowith American College of Cardiology
Susan Stimpson Novant Health, Inc.
Paul Varosy Department of Veterans Affairs
Ke Zhang M2S Inc.


Operations & Informatics Team

Lead by H.Vernon Anderson, University of Texas Health Sciences Center

First Name Last Name Organization
Carrie Bosela VQI (Vascular Quality Initiative)
Jay Crowley USDM
Denise Downing AORN
Nicholas Gawrit HEARTBASE
Roxana Mehran Mount Sinai Hospital
Rosalind Parkinson Parkinson Logistics Associates LLC
Robert Perry DoD
Michael Simanowith American College of Cardiology
James Tcheng DCRI
Ricki Wilson Vizient
Qi Zhou BCBS Maine



Organization Name
American College of Cardiology (ACC)
Aorta Medical
Association for Operating Room Nurses (AORN)
Bard Peripheral Vascular
Blue Cross Blue Shield (BCBS MA)
Cook Medical
Department of Veterans Affairs
Device Events
Department of Defense (DoD)
Duke Clinical Research Institute
FDA Patient Representative
First Databank
Franciscan Missionaries of Our Lady Health System (FMOLHS)
Hawthorne Effect, Inc.
Intermountain Healthcare
M2S Inc.
MedStreaming LLC
Mercy Health System
Mount Sinai Hospital & Health System
National Center for Health Research
Novant Health, Inc.
Parkinson Logistics Associates LLC
Society for Vascular Surgery, Vascular Quality Initiative
University of Texas Health Science Center Houston
University of Michigan
University of Virginia, MITRE
University of Washington
USDM Life Sciences
Weill Cornell Medical College - New York Presbyterian Hospital