[set fair (OP)]

https://journals.sagepub.com/doi/10.1177/0300891620974755?url_ver=Z39.88-2003&rfr_id=ori%3Arid%3Acrossref.org&rfr_dat=cr_pub++0pubmed&

More than 8% of the samples from the first week in september 2019 had CoV2 antibodies. If that were representative of the whole country, then 5 million Italians would have had covid by then, 23 doublings from the first case. Obviously, from the rest of the data, the regions where the first samples were taken isn't representative of the whole country. But in some places it was well established by the beginning of september.

[Colin2B]

I’m pretty sure we’ve discussed this briefly elsewhere. There have been a number of reports and they seem to relate to the same set of tests made on samples previously collected. I need to read this to see if it's one we've looked at already or gives more info.
If all the cases were asymptomatic that raises the question of whether the virus was mutating or where the symptoms hidden in these cases, and were isolated symptomatic cases missed. Or it could be that with low levels of infection the build up was slow.
We don't seem to have similar studies from China, but even there there are indications it was around at a lower level for a while.

I'll read through

[Bored chemist]

Any paper that says "Covid was here early" has to explain why it didn't show up.
If 8% of those Italians had covid in September, how come (with no particular controls in place)  there wasn't a pile of bodies by mid October?
The disease didn't start killing until about February.


What was the virus doing for 4 or 5 months?

Isn't it vastly more likely that there's some mistake in their experiment?

[Colin2B]

Isn't it vastly more likely that there's some mistake in their experiment?

It is very likely. I did look in the original paper I saw, but there wasn’t enough detail on what tests were being done.

[alancalverd]

8% false positives is pretty good for most antibody tests.