> For example, people in the contemporary control group weren’t tested for COVID-19, so it’s possible that some of them actually had mild infections.
How doesn't this completely invalidate the title? If we aren't actually comparing between a group that provably did not have COVID and one that did, we can't possibly draw any conclusions about the effect of having COVID compared to not having it at all.
Isn't it the opposite? If there's a strong effect in "had COVID" vs "probably did not have COVID", doesn't that prove that there would be an even stronger effect in "had COVID" vs "definitely did not have COVID"?
But where did "probably did not get" come from? It is highly likely that COVID numbers are far higher than we know about due to many mild and asymptomatic cases that go undetected.
That makes the case even stronger. Not sure where you're going with this. Unless you want to imply that everyone has had COVID already, the argument works the same.
How doesn't this completely invalidate the title? If we aren't actually comparing between a group that provably did not have COVID and one that did, we can't possibly draw any conclusions about the effect of having COVID compared to not having it at all.