Thread by @phl43

I haven't even read that paper, I just had a quick look at the tables, but I don't buy it for a second. They found that 14% of samples collected in September on participants to a lung cancer screening trial had antibodies for SARS-CoV-2. Of course, it's not a random sample, 1/n

https://twitter.com/Reuters/status/1328076605370019841

but that's still huge. There has been many studies based on that kind of samples since the pandemic started and seroprevalence was usually lower even in places where hospitals were totally overwhelmed. So we'd have to believe that, by September of last year, 14% of people in 2/n

such a sample had already been infected by SARS-CoV-2 in Italy but no one noticed anything. Not only hospitals weren't full of people with pneumonia of unknown etiology, but the Italian health authorities didn't detect any clusters of pneumonia whose cause they couldn't 3/n

ascertain and therefore didn't realize a viral outbreak was underway. I think it's totally unbelievable in a rich country like Italy, where presumably the emergence of infectious disease is monitored very closely. I know that stochasticity plays a huge role at the beginning 4/n

an outbreak, so you could imagine that a few people got infected in Italy back in September, but it just petered out. However, if 14% of the people in the sample had antibodies for SARS-CoV-2 at the time, even taking into account the fact that it wasn't a random sample, 5/n

it's pretty clear that we're not talking about that kind of scenario, it would mean that the virus would have already spread a lot in the population. But if this had been the case, then surely people would have noticed the bodies piling up, so I just don't buy it. Frankly, 6/n

even if they had found just one case in September in Italy, I wouldn't buy it because I think we have very good reasons to think the virus originated from China, but regardless of whether I'm right about that I think the fact they found such a high rate of people with 7/n

antibodies for SARS-CoV-2 in samples collected in September is actually something that makes this study less, not more, plausible. I think it actually makes it totally unbelievable. I don't know what's wrong with it, whether they mixed up the samples, used bad serological 8/n

assays or whatnot, but I'm quite sure that *something* is wrong with it and hopefully we'll learn what eventually. 9/9