The Workbench · Craft
A PMA panel referral is the default, not a choice
This blog has already covered how a PMA's filing review only checks that the application is complete enough to review, and the separate question of what FDA's review timelines actually promise once that gate clears. Neither post reaches a step that a lot of PMA timelines quietly assume is optional: referral to an outside advisory panel. Section 515(c)(2) of the FD&C Act doesn't treat that referral as something FDA elects case by case. It's the default outcome for a PMA, narrowed by one specific, named exception — and a submission plan that treats a panel meeting as a possibility rather than the baseline is planning around the wrong assumption.
The statute starts from referral, not around it
Section 515(c)(2), as amended by the Safe Medical Devices Act of 1990, has FDA refer a PMA to an advisory committee for review and recommendation, and grants an applicant the right to request that referral directly. The exception runs the other way from how most sponsors assume it works: FDA can skip the panel only where it determines the application substantially duplicates information a panel has already reviewed — not simply because the device type is now familiar to the review division, and not because the sponsor would rather avoid the delay. A first-of-a-kind device, with no prior panel record to duplicate, has no path around the referral at all.
21 CFR 814.44 and Part 14 govern what happens once one is convened
Once FDA refers a PMA to a panel, 21 CFR 814.44 and the general advisory committee procedures at 21 CFR Part 14 take over. Part 14 requires public notice of the meeting in the Federal Register, and the committee's business is conducted as a public meeting under the same framework that governs FDA's other advisory committees. The panel's job at the end of it is a final report to FDA carrying its recommendation and the basis for that recommendation — the regulation is specific that this report can take the form of a signed meeting transcript, not necessarily a separately drafted document, which is a smaller administrative artifact than the phrase “final report” tends to suggest.
The panel's recommendation doesn't bind the decision
Nothing in 515(c)(2) or 814.44 makes the panel's recommendation FDA's decision. FDA reviews the PMA together with the committee's report and recommendation and then issues its own order — an approval, an approvable letter, a not-approvable letter, or a denial — on its own authority. In practice FDA's decisions align with panel recommendations far more often than not, but the legal structure is advisory review feeding an agency decision, not a vote the agency is bound to certify. A sponsor's internal tracker that logs a favorable panel vote as the outcome, rather than as an input FDA still has to act on, has recorded the wrong event as the milestone.
The Federal Register notice sets a planning date of its own
Part 14's public-meeting requirement isn't a formality that happens after a referral decision is made and forgotten — it puts a real date on the calendar. FDA has to publish notice of the meeting in the Federal Register at least 15 days ahead under the statute's own minimum, though in practice CDRH typically publishes closer to 30 to 60 days out to give panel members, the public, and the sponsor itself time to prepare. A sponsor's own submission timeline should treat that notice date as the earliest real signal that a panel meeting is actually being scheduled, rather than waiting on an informal signal from the review division that a referral is coming.
The filing clock keeps running underneath the panel process
21 CFR 814.44(c) ties FDA's ultimate decision to the later of 180 days from the PMA's filing date under 814.42, or the date determined under 814.37(c) where a major amendment has extended that filing clock. That 180-day figure isn't suspended while a panel meeting is arranged, noticed, and held — the referral happens inside the same review period the sponsor is already tracking, not on a separate calendar bolted onto the end of it. A submission timeline that adds panel review as extra time after the ordinary review clock has misunderstood where the meeting actually sits.
Where this meets the file
A PMA tracker needs a field asking the 515(c)(2) question directly — does this application substantially duplicate a record a panel has already reviewed, or is a referral the applicable default — rather than treating panel review as a contingency to plan for only once FDA raises it. A PMA readiness worksheet built around that question is previewed in the launch catalog. If your program has been through a panel referral and tracks it differently, the shelf takes that correction directly.
The Regulatory Toolkit launches soon — a free shelf of source-mapped templates, checklists and browser-only tools for regulatory teams. Get one email when it opens, or contribute a template.