One in five early US coronavirus tests had nothing to do with travel

In December 2019 a cluster of pneumonia cases appeared in Wuhan. By **4
February 2020** there were about 20,000 confirmed cases in China and **159
in 23 other countries, including 11 in the United States.**

The United States screened travellers from Wuhan starting 17 January, and
CDC activated its Emergency Operations Center on 21 January with a **call
centre staffed by physicians and nurses 24 hours a day.** Through **31
January** it responded to inquiries about approximately 650 people and
tested 210.

Who was tested

The 20% is the finding. A screening system designed around arriving travellers
finds arriving travellers. One in five people tested had not travelled at all —
they had been near somebody in the United States — which is what the start of
domestic transmission looks like from inside a surveillance system still
pointed outward.

Hence the recommendation, aimed squarely at clinicians: **remain vigilant
regarding possible exposure not only among returning travellers, but also among
persons in close contact with patients in the United States**, and **consult
local and state health departments** for both.

The constraint behind the numbers

**Public health laboratories were not yet conducting testing during the
period covered by this report, while awaiting Food and Drug Administration
emergency use authorization for the test.**

All testing was conducted at CDC. Authorisation came on 4 February.

That single fact explains the shape of everything else. When one laboratory in
the country can run the test, a criterion for who qualifies is not a clinical
judgement but a rationing rule — and 210 tests in two weeks is what a national
capacity of one looks like.

The criteria, and the room left in them

To be a "person under investigation" during 17–31 January required **fever and
symptoms of lower respiratory tract illness** — cough or difficulty breathing —
plus an epidemiological risk: **travel from Wuhan, close contact with a
confirmed patient, or close contact with an ill person under investigation.**

And then, deliberately:

**Given the evolving understanding of 2019-nCoV epidemiology, testing was
recommended for some persons who did not strictly meet the PUI definition,
based on clinical discretion.**

Which is the right answer for a disease nobody yet understood, and also the
only reason any of the contact-risk cases were tested at all.

Source: Centers for Disease Control and Prevention, MMWR.