Monday 4 May 2020

Smoking and COVID-19: evidence update


A UK study in Lancet Infectious Diseases found that smokers were half as likely to test positive for the coronavirus. The unadjusted odds ratio was 0.59. After adjustments it was 0.49. This is important because it suggests that smokers are under-represented in coronavirus wards because they are less likely to contract the disease in the first place (and not necessarily because they are less likely to develop severe symptoms). A French prevalence study published last month suggested the same thing.

Chinese study of 202 COVID-19 patients finds that only 7.9 per cent had any smoking history. The authors don't comment on this finding. (The Chinese smoking rate is 27 per cent.)

Study from Kuwait finds that only 4 per cent of COVID-19 patients were smokers, but finds that smokers were more likely to progress to intensive care and die.

Swiss study finds only 4.5% of COVID-19 patients are active smokers. The authors don't mention this in the text.

Study from Mexico finds a smoking prevalence among patients of 9%. If admitted, smokers were somewhat more likely to need ICU and less likely to survive.

Spanish study based on an online survey finds that vapers and tobacco users are half as likely to get the coronavirus. Odds ratios are (0.45 95% CI 0.28-0.71) and (0.52 95% CI 0.27-0.98) respectively. Not the strongest study but in line with the first study in this list and the first to look at e-cigarettes.


Two meta-analyses have been published in the last week. Farsalinos et al. suggests that smokers are 78 per cent less likely to ‘have an adverse outcome’ from COVID-19. Gonzalez-Rubio et al. suggests that smokers are 82 per cent less likely to be hospitalised with COVID-19.

Two new studies came from Italy. Colombi et al. found that only 2.5% of COVID-19 patients were current smokers. However, the same table appears to show that 7.6% were smokers. The authors don't produce odds ratios or comment on the smoking findings.

The other Italian study (Gaibazzi et al.) is titled 'Smoking Prevalence is Low in Symptomatic Patients Admitted for COVID-19.' It finds that only 4.8% of COVID-19 patients were smokers, despite the national smoking rate being 24%.

A large study from the UK (Williamson et al.) found that only 6.9 per cent of the people who died from COVID-19 were smokers. This translates into a 20% reduction in risk, a 20% increase in risk or no significant effect at all, depending on how the figures are adjusted for other factors.

Finally, a study from New York published last month that I had previously missed found that only 5.1% of COVID-19 patients were smokers. The authors do not comment on this.


ASH reckons that 300,000 people have given up smoking as a result of COVID-19 fears. I share Taking Liberties' suspicions about this figure. It is derived from a survey of 1,004 people which found that two per cent of smokers had quit in recent weeks. The poll isn't available online at the moment, but if it was representative of the UK population - as it is obviously supposed to be - there would be about 150 smokers in it. If two per cent of them quit, that is three people. It seems bold to take such tiny numbers and translate them into the claim that 300,000 smokers have quit.

Be that as it may, ASH say that there has never been a better time to quit. Given how few smokers are being hospitalised with COVID-19, one could argue the opposite. If the evidence continues to show smokers under-represented in coronavirus wards, the hashtag #QUITFORCOVID could take on a sinister double meaning.

How is the evidence looking? Last week didn't see as many new studies as the previous week, but there were a few worth mentioning.

The NEJM published a study of 8,910 COVID patients from '169 hospitals in Asia, Europe, and North America'. Unusually, it found that white people were more likely to die from the disease once admitted to hospital than other ethnic groups. It also found that smokers were more likely to die once admitted. However, although the authors didn't comment on it, it also showed that there were fewer smokers admitted to hospital than you would expect. Just 5.5% were current smokers, well below the average in almost any country.

After finding that only 1.3% of patients were smokers in its last report, the US Centers for Disease Control have stopped publishing data on smoking - which is, er, odd. However, a few states are still doing so. In Georgia, the rate is just 5.2%. In Oregon, albeit from a smallish sample, the rate is 1.4%. There must be some more statewide data. Let me know if you find it.

A large study of 16,749 COVID patients in the UK was published last week. The authors don't discuss smoking specifically, but one of their graphics suggests that the smoking rate was around five per cent. The smoking rate in the UK is around 15 per cent.

I have my doubts about this chart. The study is based on a massive dataset from ISARIC which holds information on 19,809 COVID patients, of whom 514 were smokers. That is a tiny proportion, but smoking status is only established for 5,924 of them so a better estimate is that 8.7% of the patients were smokers. That is still well below the national rate.

Also of possible interest, there was a study of 226 victims of a coronavirus (but not COVID-19) in Toronto, Canada which found that: 'Only 3.1 % of patients were current smokers'.

I'll use this blog post as a holding page for all relevant studies so, for the sake of completeness, here's the story so far:

The smoking rate of COVID-19 patients in Chinese hospitals was 10 per cent in Yang et al., 6.7% in Wan et al., 3.9% in Mo et al., 7% in Huang et al., 9% in Dong et al., 1.9% in Guan et al., 6% in Zhou et al., 1.4% in Zhang et al., 9.4% in Du et al. and in 6.4% in Liu et al. In Shi et al., only 8.2% of cases had any smoking history. The smoking rate in China is 27%.

The highest smoking rate recorded in any of the studies is 18.5% in Kim et al.'s study from South Korea and that is still well below the national smoking rate. In New York, it is 6.7%.

This study from the USA looked at 3,789 US military veterans aged between 54 and 75 who were tested for COVID-19, of whom 585 tested positive. Smokers were heavily under-represented with an odds ratio of 0.45 (95% CI 0.35-0.57), ie. they were 55 per cent less likely to test positive.

This study of 661 people tested for COVID-19 in a badly hit area of France reported that: 'Smoking was found to be associated with a lower risk of infection (7.2% versus 28.0% for smokers and non-smokers, respectively), and this association remained significant after adjustment for age (OR = 0.23; 95% CI = 0.09 –0.59) or occupation (OR = 0.27; 95% CI = 0.10 –0.71).'

This suggested that smokers were 70-80 per cent less likely to test positive for COVID-19, in line with a previous study from France.

Finally, there is some evidence that smokers are more likely to suffer complications and/or die once hospitalised with COVID-19. Then again, there is evidence that they don't. See this post for a discussion of that.

Let me know if I've missed any studies (note that I am not including reviews, meta-analysis or commentaries).

No comments: