Let’s start with what most everyone agrees on: The Unique Device Identification (UDI) program is a swell idea.
It gets a little trickier after that.
In extensive comments, the Advanced Medical Technology Association (AdvaMed), Boston Scientific, and Merck, among more than a dozen others, generally voice support for the UDI concept, while finding lots and lots to say about where the FDA’s September Draft Guidance could use improvement.
Noting that implementing UDI will be a “costly proposition,” AdvaMed stresses that the length and complexity of the implementation plan demands a “living document” approach that will allow industry and the FDA to update and improve the guidance as both sides learn more during set-up. AdvaMed follows with 61 specific comments, with suggested changes.
Coviden, manufacturer of medical devices and medical supplies, echoes AdvaMed’s comments, and tosses another 22 into the mix, including a request that the guidance remain open for feedback and comment until the September 24, 2014 implementation deadline.
Merck, among other commenters, requested clarification and summarization regarding the scope of products for which data must be submitted to the Global Unique Identification Database (GUDID). Merck also asked FDA to add information regarding deadlines for submitting data to GUDID.
To pick one of their examples, and joining several other commenters in making this point, Boston Scientific claims the data elements column “Required?” is unclear because it fails to clarify if it is required to follow the rule based on regulatory requirements or validation requirements. “The meaning of ‘required’ should be clarified so that BOTH regulatory and system validation requirements are clearly identified in this guidance.”
FDA’s got its work cut out for it here, particularly with the recent departure of its UDI guru, Jay Crowley, for the greener fields of consultantdom.
We can offer some small consolation though: Crowley leads a webinar on UDI implementation from his new professional perch. Information is here: