A PMA approval order can bind long after clearance

A PMA file tends to get built, and read afterward, as if the approval order is the finish line — the last document dated before the device can be sold. 21 CFR 814.82 treats the order as something closer to a starting line for a second, quieter set of obligations. It lets FDA attach postapproval requirements to the order itself, as a condition of approval, and those requirements can run for as long as the device stays on the market — not until the file closes, because under this section the file never really closes.

The order can restrict the device, or commit it to being watched

Section 814.82(a) names what a postapproval requirement can be: a restriction on the sale, distribution, or use of the device, or continuing evaluation and periodic reporting on the device's safety, effectiveness, and reliability for its intended use. Neither is boilerplate FDA attaches by default — the section requires FDA to state, in the approval order itself, the reason or purpose for the requirement, the number of patients to be evaluated, and the reports that have to be submitted. That makes the approval order an operative instruction set a quality file has to keep reading long after clearance, not a letter that gets filed once and archived.

Annual reporting runs on a calendar, not on events

Section 814.84 sets the routine baseline underneath any device-specific condition: continued approval of the PMA is contingent on submitting postapproval, or annual, reports at intervals of one year from the date of approval, containing the information 814.84(b) requires. That clock runs from the approval date regardless of what has or hasn't happened since — a device with a quiet year still owes the report on schedule. It's a different shape of obligation than the event-triggered reporting this blog has already covered on the adverse-event side: an MDR fires because something happened; a postapproval annual report fires because a year passed, whether or not anything did.

A restriction is a condition of approval, not a labeling suggestion

Where an order restricts the device's sale, distribution, or use, that restriction is a term of the approval itself — not a recommendation a distributor is free to read as advisory or a limit the labeling alone is expected to enforce. A restriction on which facilities can purchase the device, or which patients it can be used on, sits on the same footing as the reporting duty: both are conditions 814.82(a) authorizes FDA to impose, and both carry the same consequence for a manufacturer that lets one slide because it never generated a report to miss.

A named postapproval study is a heavier, individually negotiated condition

Not every PMA carries one, but where an approval order requires a dedicated postapproval study, it's a materially bigger commitment than the routine annual report folding in whatever safety data accumulated on its own. A required study specifies its own patient count, its own follow-up duration, and its own protocol — and it has to be tracked as an open, individually scoped commitment against the terms the order actually states, not absorbed into the general reporting cadence as if it were one more line item. A quality file that can't point to the specific order language a study answers to has lost the thing that makes the commitment enforceable in the first place.

Noncompliance reaches back to the approval itself

814.82's own text closes the loop plainly: failure to comply with any postapproval requirement is a ground for withdrawal of approval of the PMA. That's not a separate administrative penalty sitting beside the clearance — it's the same approval the requirement is a condition of, and a manufacturer that treats postapproval reporting as a lower-priority obligation than the original submission is misjudging where the actual leverage sits.

Where this meets the file

A postapproval tracking record that reads straight off the approval order's own conditions — restrictions, the annual-report clock, any named study, each with its own owner and due date — rather than a generic renewal calendar built once and reused across PMAs, is previewed in the launch catalog, alongside the premarket case a PMA has to make in the first place, covered in this blog's note on De Novo's own benefit-risk case. If your program tracks postapproval conditions differently, the shelf takes that correction directly.