clinicians.build · interactive · august 30, 2026

Frozen at the Edge

From out there you look like you've stopped. Your watch says otherwise. Here is what waiting actually looks like when you draw all of it at once: 69,975 FDA device clearances, one row per year, stacked by how long each one took to cross.

Data: FDA 510(k) Premarket Notification database, all decisions 1990–2024 (release of August 17, 2026)
Frame: Carlo Rovelli, White Holes — the horizon, and the ship that vanishes from the shore

If you fall toward a black hole and I stand out here and watch, I see your clock slow until you stop at the edge and hang there. You never cross. You, meanwhile, notice nothing — your watch keeps ordinary time.

Both readings are correct. Neither one is the real one.

That is every clinician who has ever shipped a working thing in a weekend and then watched it sit in review. So it is worth looking at a horizon we can actually measure, from the outside, for thirty-five years.

Thirty-five years of falling

Each row is one year of FDA 510(k) decisions. Moving right is a longer wait: the ridge shows what share of that year's clearances took each span of calendar days from receipt to decision. The red dot is the median — the middle submission. The grey line at 180 days is roughly where a wait stops feeling like a queue and starts feeling like a wall.

Hover or tap any year — press replay to watch it build
Median thread 180-day line p90 thread
Year
decision cohort
Clearances
510(k) decisions
Median wait
calendar days
One in ten
waited longer than
Over 180 days
share of that year
median share of that year's clearances 180 days last bin pools everything over 540 days · hover a row, tap on mobile

Three things this picture says

The horizon moves. In 2005 the median 510(k) crossed in 59 days and only 11.8% took longer than six months. In 2024 the median was 132 days and 37.2% took longer than six months. Same statute, same submission type, more than double the outside-clock wait.

The horizon has moved before and come back. The bulge at the top of the chart is real and it is worse than anything since: 1993's median was 176 days, its 90th percentile 399, and nearly half of all clearances that year took more than six months. Then it collapsed. Twelve years later the same agency was clearing devices faster than it ever had. If you had been inside that 1993 queue, every reasonable person watching from the shore would have told you the project was dead.

The tail is where the fear lives. The median moved by 73 days between 2005 and 2024. The 90th percentile moved by 78, and in 2022 it hit 409 days. Almost nobody plans around a median. People plan around the story they heard about the submission that took fourteen months.

A horizon isn't a wall. It's two clocks disagreeing — and the reading on either one tells you as much about where the observer is standing as about the thing being observed.

The clock this picture cannot see

Now the part that should make you hold all of the above loosely.

FDA does not grade itself on the number drawn above. Its performance goal under MDUFA V is a decision within 90 FDA Days for 95% of standard 510(k)s — and FDA Days count only the time the submission is on FDA's desk. When the agency issues an Additional Information request, the clock stops, and restarts when the sponsor answers. Months of a submission's calendar life can belong entirely to the company that filed it.

The public database publishes a receipt date and a decision date and nothing in between. So every ridge above is the outside clock by construction, and the drift it shows is the sum of two things it cannot separate: FDA review time, and sponsor hold time.

Which is the whole point, and it survives the caveat. The outside clock is not lying to you. It is measuring something real, and the thing it measures is not whether you are moving.

Where this picture is thin

⚠︎ AI-generated · not reviewed by a human · verify against the linked sources before relying on it. Every median, percentile and share on this page is this page's own arithmetic on the public FDA 510(k) file — not an FDA analysis, not an MDUFA performance figure, and not regulatory advice about any submission.