Plague outbreak data: what the record shows

Data on the 2026 plague reports: daily reporting volume, the 41-outlet source pool by region and language, official statements, and how many claims the record can settle.

Compiled from the same dataset as every other page on this site · refreshed 7 Oct 2026, 18:00 UTC · sources and method

The figures we track

As of 7 Oct 2026, 18:00 UTC

Five numbers describe this event in the public record. None of them is a case curve: the underlying record is a set of dated statements, not a surveillance feed, so we publish each figure with the source that stated it and the date we recorded it. Nothing here is interpolated or modelled.

1 Deaths under investigation

A 28-year-old laboratory worker at the Irkutsk anti-plague institute died in Shelekhov. Authorities have not confirmed the cause of death.

Recorded 2026-10-07

2 Suspected cases reported

The first is the worker who died, whose cause of death has not been officially confirmed. WHO has asked Russia to clarify media reports of a second employee with pneumonia of undetermined cause.

Recorded 2026-10-07

0 Laboratory-confirmed cases

No case of plague has been confirmed by laboratory testing in this event as of the latest official statements.

Recorded 2026-10-07

~200 Contacts under medical observation

Authorities isolated about 200 contacts of the worker and introduced quarantine restrictions at the hospital where she had been treated. Rospotrebnadzor later said more than 90 percent of identified…

Recorded 2026-10-07

~5,000 Laboratory tests reported

Rospotrebnadzor said the results of almost 5,000 laboratory tests did not reveal dangerous pathogenic microorganisms, and that the woman had been vaccinated against the infections she worked with.

Recorded 2026-10-07

Suspected and monitored figures are drawn from media and official reporting and change frequently; each figure below links to its source.

Reporting volume, day by day

6 days · peak 22

How much was published about this event each day, across every outlet we aggregate. This is the honest proxy for how fast the story is moving — we can measure what was published, not how many people are infected. Dated events from our timeline are shown alongside.

1 10-02 ◆2

2 10-03 ◆2

3 10-04 ◆1

9 10-05 ◆2

12 10-06 ◆3

22 10-07 ◆8

Bar height is items published that day; ◆N is how many dated events fall on it. 4 further items in the dataset are background explainers published before the event window and are excluded from this chart. The same events in full are on the timeline.

Official statements on the record

18 dated events

The spine of the record is what named bodies said, and when. We label each entry by how well it is established: confirmed means an official body said it or two independent outlets report it; reported means one outlet reports it; disputed means sources conflict.

8Confirmed

10Reported

0Disputed

10-02 2

10-03 2

10-04 1

10-05 2

10-06 3

10-07 8

Each row is one day; segment widths are proportional to how many of that day's events are confirmed, reported or disputed. Full detail on the timeline.

Who is reporting this

41 outlets

The dataset behind this site: 53 items from 41 outlets in 15 languages. If a story is only covered in one country or one language, that is worth knowing — so this is published rather than summarised.

By region

Europe 17 · 32%
Russia 16 · 30%
Asia 6 · 11%
International 6 · 11%
China 3 · 5.7%
Mongolia 3 · 5.7%
Americas 1 · 1.9%
Oceania 1 · 1.9%

By language

English 16 · 30%
Russian 16 · 30%
Bulgarian 3 · 5.7%
German 3 · 5.7%
Chinese 3 · 5.7%
Spanish 2 · 3.8%
Ukrainian 2 · 3.8%
Arabic 1 · 1.9%
Danish 1 · 1.9%
French 1 · 1.9%
Japanese 1 · 1.9%
Korean 1 · 1.9%
Mongolian 1 · 1.9%
or 1 · 1.9%
Swedish 1 · 1.9%

By kind of item

Reporting 41 · 77%
Analysis 5 · 9.4%
Explainers 4 · 7.5%
Official statements 3 · 5.7%

Outlets following the story

Echo (ЭХО) 5
Meduza 4
Bulgarian Telegraph Agency (BTA) 2
NEWS.ru 2
RTVE 2
Ukrinform 2
World Health Organization 2

34 of 41 outlets have published a single item we could reach; only outlets with more than one are charted, because a list of singletons is not a signal.

How much of this is settled

10 claims checked

We check the claims people are repeating. The most useful number about a story like this is not how many are true but how many are still open: 5 of 10 (50%) cannot be settled on the public evidence either way.

False 3 · 30%
Mixed 2 · 20%
No public evidence 2 · 20%
True 2 · 20%
Unverified 1 · 10%

A board that only ever said "false" would not be a check, it would be a nudge. Ours carries claims we rate true, claims we rate false, and claims the record cannot settle — and the build fails if every claim ends up with the same verdict.

Read the checks

Where these numbers come from

Derived, not estimated. Every figure on this page is computed from the records in our dataset: 53 aggregated articles, 18 dated timeline events, 10 claim checks and the official statements we quote. Nothing is interpolated, modelled or estimated, and no number appears here that is not also traceable to a named source elsewhere on the site.

What this page deliberately does not show. There is no case curve, because the record is a set of dated statements rather than a surveillance feed — a line drawn through one or two real points would be invented. There is no per-outlet rating, because we record what outlets published rather than scoring them. And there is no forecast of any kind.

How it updates. The aggregates are rebuilt from the dataset every time the site is built, so this page cannot drift from the articles it summarises. The figure records accumulate: each time an official number changes, the new value is appended with its source and date rather than overwriting the old one.

Provenance

Dataset built: 53 items
Outlets: 41
Languages: 15
Event window: 2026-10-02 → 2026-10-07
Figures tracked: 5

Aggregates rebuilt 7 Oct 2026, 18:00 UTC