DrugDashboards

FDA plant inspections, by country

Every drug sold in the US is made in a plant the FDA can inspect, wherever in the world it sits. When inspectors find serious problems they classify the visit OAI — official action indicated — the strongest public warning a facility gets. This page asks a simple question: what share of the plants in each country came out of inspection with that flag?

ⓘ How this was built
Data sources
  • FDA Data Dashboard — inspection classifications (drug and biologic program areas)
  • FDA DECRS — Drug Establishments Current Registration Site (facility country)
  • DailyMed SPL drug labels (which plant makes each NDC, by DUNS)
  • FDA drug shortages via openFDA
Method

Counted per facility, not per inspection: a plant inspected four times would otherwise count four times. A country's rate is the share of its FDA-inspected drug plants with at least one OAI classification in the window since Sep 4, 2021. Countries with fewer than 20 inspected plants are omitted as too small to rate. "Drugs made there" counts distinct generic names whose FDA label names a manufacturing site in that country.

Pipeline and calculations built by Claude (Fable 5) from the public datasets above. Educational only — verify against primary sources before acting.

Inspections
17,695
Plants inspected
12,321
Plants with an OAI
807
Countries
71

Read this before the numbers

FDA does not inspect every plant on the same schedule. Foreign sites are chosen substantially by risk — history, product type, prior findings — while much US inspection volume is routine. A higher OAI rate abroad therefore reflects both how plants perform and which plants get visited, and the two can't be separated from public data. Read these as “what inspectors found where they looked,” not as a clean ranking of national manufacturing quality.

When a plant fails inspection: the drugs with the least slack

Plants whose most recent two years include an official-action (OAI) inspection, and the drugs their FDA labels name them as manufacturing — filtered to where the finding could actually bite: three or fewer active makers, or a shortage already under way. If one of these plants deteriorates further, these are the supplies with the thinnest cushion.

How much does that matter? Measuring the same molecules in periods with no preceding inspection failure: 14.4% went into shortage within 12 months of an OAI, against 6.7% otherwise — about 2.13× the risk. That is an association, not proof of cause: FDA targets inspections by risk, so the same underlying trouble may drive both.

Manufacturing sites come from public FDA (SPL) labels, which name an establishment for roughly a quarter of marketed NDCs — a plant missing here may simply not be named on a public label. Maker counts are active labelers in the trailing 30 days.

Share of inspected plants with an official-action finding

Last five years · countries with at least 20 inspected drug plants · bar length is the OAI share

CountryPlantsOAI shareSome findingUS drugs made there
United States8,567
6.7%
37%1,360
India880
10.3%
66%1,057
China457
7.0%
61%272
Germany245
0.8%
73%175
Canada201
7.5%
52%199
Italy194
2.1%
81%135
France176
3.4%
76%67
Spain148
2.0%
68%71
United Kingdom146
4.1%
71%55
Japan139
2.9%
65%25
Poland116
0.9%
19%10
Switzerland88
1.1%
68%102
Ireland64
0.0%
91%91
Netherlands56
1.8%
59%14
Mexico55
14.5%
58%9
Taiwansmall sample46
6.5%
70%55
Belgiumsmall sample46
0.0%
83%49
Australiasmall sample45
11.1%
40%16
Bulgariasmall sample42
0.0%
21%—
Denmarksmall sample39
0.0%
69%9
Argentinasmall sample33
6.1%
30%—
Czechiasmall sample33
0.0%
55%3
Brazilsmall sample28
3.6%
46%2
Singaporesmall sample25
12.0%
80%19
Türkiyesmall sample24
20.8%
50%11
Swedensmall sample23
4.3%
87%17
South Africasmall sample23
4.3%
30%1
Austriasmall sample20
0.0%
85%29

“Some finding” combines OAI with VAI (voluntary action indicated) — a real but less serious result. For scale, the US rate is 6.7%. Countries under 50 inspected plants are marked small sample and drawn in grey: with a denominator that small a single facility swings the rate by several points, so a high number there is much weaker evidence than a lower one from a large country. Plant counts come from inspection records while the drug column comes from the separate registration and label data, so coverage differs and a country can show plants with no traceable drugs. Where drugs are made →

Recent official-action plants, by how much they make

Plants with an OAI in the last five years, ranked by the number of distinct US drugs whose labels name them as a manufacturing site. This is the blast radius: a supply problem at the top of this list reaches a lot of shelves.

An OAI is a finding about the plant, not a judgment about any specific drug it makes — and problems are often fixed. Drug counts cover products whose labels list the site for a manufacturing step.

Educational only — not clinical or purchasing advice. Inspection classifications are published by FDA and can be revised.