PredicateScope

What a 510(k) predicate is, and what this page can say about one

A 510(k) submission reaches the US market by demonstrating that a device is substantially equivalent to one already legally marketed. That earlier device is the predicate. The comparison is argued by the submitter, in a submission, and reviewed by FDA.

FDA publishes the outcome of that process and not its reasoning. The openFDA 510(k) endpoint carries the applicant, the device name, the product code, the review panel, the decision and its date, the receipt date, the regulation number, and two review flags. It carries no predicate field. The predicate citation lives inside the 510(k) summary document, and a cleared submission may file a statement instead of a summary, so no complete public dataset of predicate relationships exists to draw a lineage graph from. This tool therefore maps what the API returns as data — every clearance under a product code, and how long each one took — and links the FDA summary document when the record says one was filed.

How the review interval on this page is computed

Two published fields, one subtraction: decision_date minus date_received, in UTC, reported as calendar days. That is the whole method, and it is the reason the number can be checked against the record beside it.

It measures elapsed time between two FDA milestones. It does not measure FDA review time.

The span includes every day a submission sat on hold while the submitter prepared a response to an Additional Information request, plus administrative handling at both ends. Two submissions with identical FDA review effort can differ by months on this measure because one answered a deficiency letter in a week and the other took a quarter. It is not a MDUFA decision goal, it is not comparable to one, and it says nothing about any submission that has not been decided.

Records missing either date are excluded from every time figure, and the count of what was excluded is printed in the source line under the figure — never dropped. A decision year with clearances but no receipt dates still gets a row in the year table, showing n=0 and no median, because a missing row and a zero row mean different things.

What this tool does not do

• It draws no predicate lineage graph, for the reason above: the data does not exist publicly.
• It does not assess substantial equivalence, and it identifies no clearance as suitable for any submission.
• It ranks nothing, scores nothing, and generates no text about any record.
• It says nothing about the outcome or duration of a submission that has not been decided.

Choosing a predicate is a judgement made by the people preparing a submission, against the device in front of them. A tool that made that call from six public fields would be making it up. This one presents the record and stops.

Why we are giving it away

Because it is public data and a tedious job, and because the honest way to show you how we work is to hand you something that works rather than describe it. If it saves your team a day, that tells you more about whether to talk to us than any page on this site.

The same discipline that shows up here — every figure carries the query it came from, and a field FDA did not populate says so instead of showing a zero — is the one that shows up in the DHF gap reports we charge for.

About openFDA lag

openFDA mirrors FDA source systems on its own schedule and every response states when it was last refreshed. That date — not today's date — is what this page stamps on every figure, in every source line, and in the disclosure block: run a search and the Data date cell above the results is the one openFDA returned, whatever the calendar says. If a response ever carries no date at all, this page withholds the figures rather than dating them itself. The exact request URL behind whatever is on screen is printed under it, so any figure here can be reproduced against the same endpoint.

Questions

What is a 510(k) predicate device?

A 510(k) submission clears a device by demonstrating it is substantially equivalent to a device already legally marketed in the United States. That earlier device is the predicate. The comparison is made by the submitter and reviewed by FDA; nothing about it is automatic.

Does PredicateScope show which device each clearance cited as its predicate?

No. openFDA publishes no predicate field. The citation appears inside the 510(k) summary document, and a cleared submission may file a statement instead of a summary, so no complete public dataset of predicate relationships exists. This tool shows what the API returns as structured data and links the FDA summary document when the record says one was filed.

What do the review days on this page measure?

Calendar days from date_received to decision_date, both published fields on the same record. That span is elapsed time between two FDA milestones. It includes the days a submission spent on hold while the submitter answered an Additional Information request, and administrative time. It is not FDA review-clock time and it is not a MDUFA decision goal.

What happens to records missing one of those dates?

They are excluded from every time figure and counted in the source line under it, so a panel reads n=142 with 8 records excluded rather than quietly reporting 142 as though it were 150. A year with clearances but no receipt dates still appears in the year table, with n=0 and no median.

How current is the data?

Every openFDA response carries a last_updated date, and that is the date printed on every figure and in the disclosure block on this page. openFDA lags FDA source systems, so that date is normally behind today. This page never stamps a figure with the date you happen to be reading it.

Do I need an account or an API key?

No. The page queries the public openFDA endpoint straight from your browser with no key and no sign-in, and stores nothing about you. The exact request URL is printed under every result, so you can run it yourself and get the same records.