A third party reviews a 510(k); FDA still decides

A sponsor who routes a 510(k) through FDA's Third Party Review Program is often chasing one thing: speed. An accredited reviewer, paid directly by the sponsor, works the file instead of waiting in CDRH's own queue. That's a real advantage, and it's easy to read the program as outsourcing the clearance decision itself along with the review. Section 523 of the Food, Drug, and Cosmetic Act doesn't outsource that decision. It authorizes FDA to recognize qualified third parties to conduct the review and forward a recommendation — the word the statute actually uses — and it leaves the clearance determination exactly where every other 510(k) leaves it: with FDA.

Recognition runs through 523(b), eligibility through FDA's own list

Section 523(b) is what lets FDA accredit an organization as a third party review organization in the first place — a recognition FDA can suspend or withdraw if the organization stops meeting the section's requirements or starts posing a risk to public health. That accreditation authorizes an organization to do the work; it doesn't hand that organization authority over which devices it can review. That second question runs through a separate mechanism: FDA maintains and periodically updates a published list of eligible device types, limited to low-to-moderate-risk Class I and Class II devices, and a device carrying a novel intended use, a significant software component, an open cybersecurity question, or any other unresolved safety concern generally doesn't make the list regardless of its classification. A sponsor checking whether a device qualifies for third-party review is checking that list, not guessing from the device's class alone.

What actually crosses the reviewer's desk

Once a sponsor selects an accredited organization from FDA's roster, that organization does the primary substantive work a 510(k) review normally does inside CDRH — reading the submission against the predicate comparison, the performance data, the labeling, and issuing deficiency questions back to the sponsor the same way an FDA reviewer would. Where the process diverges is what happens once that work is done. The reviewer doesn't issue a clearance letter. It forwards the submission, along with its own documented review and its recommendation, to FDA — and FDA is the one that reads that package and decides.

The recommendation is evidence, not a decision already made

FDA is explicit that third party review organizations don't have authority to clear a device, and that FDA retains full authority over the clearance determination once a recommendation reaches it. That's not a formality tacked onto an otherwise-finished process, in the way a rubber stamp might be. FDA's own guidance describes the agency completing its review of the recommendation and issuing its decision within a set window after receiving it — a real, bounded review, not an open-ended second opinion FDA might get around to. A sponsor that treats the third party's recommendation as the substance and FDA's step as paperwork has the two roles backwards: the accredited reviewer produces the analysis, and FDA is still the one weighing it against the substantial equivalence standard before anything clears.

The same shape shows up elsewhere in the file

A Refuse-to-Accept review checks a 510(k)'s completeness before anyone reads its substance, and a PMA's own filing review does the same at a different tier — both are gates a submission passes through on the way to a decision, not the decision itself. Third party review sits in a similar structural spot, one step further along: a full substantive review has actually happened by the time FDA sees the file, but the review and the decision are still two separate acts performed by two separate parties, and a submission tracker that logs only “third party review complete” has recorded the wrong milestone as the finish line.

Where this meets the file

A submission tracker built around this program needs a field the ordinary 510(k) timeline doesn't: the date the third party forwarded its recommendation, kept distinct from the date FDA actually issues a clearance decision on it, since the predicate and technological-characteristics questions a reviewer raises along the way still have to be tracked against FDA's own eventual determination, not settled by the recommendation that precedes it. A third-party-review tracking worksheet that keeps those two dates as separate, linked fields is previewed in the launch catalog. If your program tracks this pathway differently, the shelf takes that correction directly.