Wednesday 6 May 2020

The dog that didn't bark

Here's a novelty. In this study, non-smoking is listed as a co-morbidity.

Why? Because it is a study of coronavirus patients in the New York City area and the last study from NYC showed that non-smokers were more likely to be hospitalised with COVID-19 than smokers.

The smoking rate in New York is 14 per cent, and 19 per cent of Americans are ex-smokers (this figure rate is probably higher in NYC). You might therefore expect at least 33 per cent of patients in this study to be ever-smokers. Indeed, given smokers' increased risk of underlying health conditions, you might expect that number to be a good deal larger.

In fact, only 16 per cent of the patients were ever-smokers. On the face of it, ever-smokers are around half as likely to be hospitalised with COVID-19 than never-smokers.

What do the authors of the study make of this? Nothing. Smoking is never mentioned, except in the table, and there is no attempt to produce risk ratios.

I suspect that the authors ran the numbers and didn't like what they found. I also suspect that they had the figures for current and former smokers separately but combined them (as the authors of the previous NYC study did).

We're not going to get anywhere if we leave stones unturned. Maybe there is nothing in the smoking/nicotine hypothesis - I certainly have my doubts - but we can't just ignore evidence we don't like. As Dr Ruth Propper says in the current issue of Nicotine and Tobacco Research...

“Whether we approve of the results from a societal or public policy perspective is beside the point.”

In science, that should really go without saying. 

No comments: