2/n Paper is here, it's really your bog-standard epidemiological study. The authors took a large database of people who had tested positive to COVID-19 and compared them to people who hadn't in Israelhttps://www.sciencedirect.com/science/article/pii/S221321982031240X …
-
-
Show this thread
-
3/n The two groups were different in loads of ways. However, after putting all the variables into a statistical model, they found that asthma appeared to still be significantly connected to risk of COVID-19!pic.twitter.com/gci3cfcuF5
Show this thread -
4/n They then went on to suggest that this statistically significant result may mean that asthma is protective against COVID-19, although the mechanism is unknownpic.twitter.com/NXSyAXXyYU
Show this thread -
5/n Now, most epidemiologists are probably already groaning, because this is a very beautiful example of the Table 2 fallacy, which is a pretty common problempic.twitter.com/Gtus24RHP2
Show this thread -
6/n The basic issue here is that you have to be VERY CAREFUL with your statistical models. If you just lump everything in together it can make the results a bit meaningless
Show this thread -
7/n There's also a problem in this study that boils down to selection bias You see, the study wasn't looking at every person infected with COVID-19 It was looking at everyone who TESTED POSITIVE TO COVID-19
Show this thread -
8/n This creates an issue known as selection bias. The group of people that is selected to be studied (due to who got a test for COVID-19 in Israel earlier this year) may not be representative of the population in crucial ways that could have biased the results
Show this thread -
9/n For example, we know that asthma diagnoses are very variable by many characteristics, which may also make people more or less likely to go get tested for COVID-19 (young men, for example - lots of asthma, might avoid testing)pic.twitter.com/VvkzTdo93f
Show this thread -
10/n It also makes correcting for confounding challenging, because we do not know which biases are influencing the SELECTION of participants
Show this thread -
11/n Ultimately, it's a bit hard to say more than that people with asthma were less likely to have a positive COVID-19 test in the group of people presenting for testing in Israel earlier this year
Show this thread -
12/n While that's a bit less sensational than the headline, it's also a bit more truepic.twitter.com/LLhY2Dw0EK
Show this thread -
13/n I should also note - the study itself is fine, generally, but the limitations section doesn't really discuss some of the biggest drawbacks of their methodology which I don't lovepic.twitter.com/iHFE2FPRyA
Show this thread -
14/n Anyway, the most likely explanation to me for fewer people with asthma getting COVID-19 is that we told them to be careful not to catch the respiratory infection because they could easily die and so they were more careful Inhalers? Ehhhhh
Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.