(https://twitter.com/DrJBhattacharya/status/16649160363518525...)
> It assumes that if a kid gets a fever first and then an adult in the house 1 to 7 days later, then the kid got covid and passed it to the adult.
This is a bad way to infer the direction of transmission. It's possible that parents had a fever, but didn't bother to check themselves until after their kid developed a fever and they checked them.
The paper does not even test a single participant for covid, so the fever might not represent covid at all. Nor does it check for other contacts the adult or kid may have had outside the house
Additionally, the sample is self-selected: it consists of people who opt into a fever reporting app -- not representative of the population as a whole.
The result is at odds with more careful contact tracing studies, like the Icelandic study from early 2020, which concludes the opposite. The paper is not a reliable guide to the relative risk that kids pose for disease spreading.
More broadly, the paper has no policy import. Would closing schools have protected anyone from getting covid? The evidence from studies comparing places that closed them vs. places that opened them suggests no. I
That closing schools is bad for the health of children is out of the scope of the ideas considered by this study.