FDA Warning Letter Explorer

Loading data status.

FDA Warning Letter Explorer: analysis

… letters in the saved dataset.

Summary

The whole record at a glance. Set a filter on the left and your slice draws over it, then pick a tab to go deeper.

Letters per month

Warning letters issued per month to drug, device and biologics makers, 2021 to 2026. Peak: . Dashed line: 12-month average, now a month. Full months only. Hover the chart for any month.

Top 3 trends

The sections whose citation count moved the most.

Top 5 citations

The sections of 21 CFR cited in the most letters. Click one to filter.

Top 5 issue types

What went wrong, as read from the findings. Click one to filter.

Quality terms in the letters

The bigger the word, the more letters in your filter use it. Click words to filter; picking several finds letters that mention all of them.

What the agency keeps finding

Citations are pulled from the letter text. When a filter is on, the dim bar behind each value is the unfiltered total.

Trends

Letters citing each section of 21 CFR in a stated finding, by year. Change compares two periods of your choosing; both counts are shown in every row. Click a row to filter.

Compare
Sort

Most cited sections of 21 CFR

Letters citing each section. Click a bar to filter; "Details" opens the regulation and its letters.

?

The findings, in FDA's own words

The most common violation statement for each section, with how many letters state it. Click a statement to read it in full; click the section number to filter.

What went wrong, by issue type

Model-assisted issue types group the letters by quality topic. Supporting FDA passages appear in letter details where available. A letter counts once per type. Click a bar to filter.

Go deeper

The same filter from more angles.

The signal was always there.

Botable builds custom AI for regulated operations, grounded in your controlled documents. This dataset is the same kind of reading, applied to the agency's own record.

Method

This snapshot combines FDA index fields, rule-extracted citations and findings, and model-assisted issue types, requests and prose findings. Top-level descriptions use source excerpts checked against the archived FDA letter, not unchecked AI summaries. These selected passages do not summarize the entire letter. Finding interpretations and categories still require source review. Response timing appears only when supported by verified response wording; a published close-out status does not establish a firm's current compliance.

Scope

Coverage is limited to the letters retained in this saved snapshot. Missing letters or close-out records should not be interpreted as a firm's current regulatory status.

Share this view

Send the same filters, search and view to a colleague.

You choose what to post or send. Sharing a link does not change access to this site.