[hwillis]

The study isn't talking about history, it's talking about current smokers and many of the comments on the paper bring up concerns with how specific that is, and how many could have quit days before going to the hospital.

I don't know where you got 50%+. The outpatient group was 5.3% current smokers vs expected (adjusted for age and sex) 26.9%. For inpatient it was 4.4% vs expected 17.9%. The fact that both groups had the same reported rate of current smokers (within experimental error) but very different expected rates says to me that you're only getting the people who are honest or simply incapable of quitting even while sick.

Also, see this concern brought up:

> Finally, and I believe this to be the most significant piece of data supporting the null hypothesis, the prevalence of never-smokers in the general population is approximately 0.75, if one subtracts the smoking incidence rate from 100. In your patient groups, non-smokers are strongly under-represented by about a factor of 2 relative to the general population, with 31% of outpatient and 32% of inpatient being labeled as never-smokers. This suggests to me that any amount of smoking actually puts one at risk for contracting COVID-19 as defined by this paper.

[tripletao]

I'm not sure I understand the quoted concern? The study they referenced[1] showed that France was ~37% never smokers average over all ages (nowhere near 75%), and their in- and out-patients were about the same fraction (Table 2) for male, a little lower for female. The big discrepancy is under-representation of current smokers, and over-representation of former smokers. But their patients are old, and I'd expect older people to have more former smokers, since they've had more time to start and stop and since the general trend in smoking is down. I don't see that broken down by age in the paper they linked though. Maybe we'd have to dig in to the raw data, or maybe it's just not available?

In any case, many other studies of COVID-19 have found similar results, and studies of different respiratory diseases have not. I'd initially just thought people were lying too, but at some point the evidence becomes overwhelming--if the protective behavior were anything but smoking, then people would have accepted it long ago.

Of course smoking is far deadlier on average than the coronavirus, per my calculation elsewhere in this thread. No one should start smoking because of this, but I do see enough evidence e.g. that a nursing home patient (who's at very high risk of death from coronavirus, and likely to die of something else before smoking-related diseases could develop) shouldn't quit. Vaping probably gets any benefit with almost none of the health risk, though that's speculative.

1. http://beh.santepubliquefrance.fr/beh/2019/15/pdf/2019_15_1....