OP-ED: Demystifying Myths About Covid and Smoking

At the start of the coronavirus pandemic, researchers stumbled upon an unexpected finding: smokers appeared to be protected from the worst effects of Covid. This “smoking paradox” was originally discovered in a study of hospital patients in China, later it was reported in studies from Italy and France.

But it turned out to be wrong, as shown by a massive study from the UK last month. Smokers were 80% more likely to be hospitalized than non-smokers. So what happened, and how did science get things so wrong?

The mathematician Pierre-Simon Laplace once said: “The more extraordinary a fact is, the stronger evidence it needs.” The American cosmologist Carl Sagan, as we know, rephrased this as follows: “Extraordinary claims require extraordinary evidence.” And let’s be honest, for smokers whose lungs are ravaged by tobacco, it is quite miraculous to get better results with respiratory disease.

Unfortunately, Outstanding Evidence is slow, complex, and kind of boring. Public attention, on the other hand, is particularly keen to capture the extraordinary

Everything is a probability

Everything is a probability, and any new information makes us update our probabilities. There is a famous example of this in statistics that was first put forward by mathematician Joseph Bertrand (I promise, I’ll get back to the smoking paradox in a moment).

Suppose you have three identical boxes. One contains two gold coins, one contains two silver coins and the last one contains one gold and one silver coin. Randomly pick one of the boxes (let’s call it Box A). What are the chances that it has the two silver coins?

Exactly a third.

Now, without looking in the box, take out a coin. If that coin is gold, then what happens to the probability that box A was the box that contained two silver coins?

It sinks to zero. New information triggered an immediate update of the probability.

Which (finally) brings me back to Covid. In January 2020 we knew little about this virus. As soon as good evidence comes in, our probabilities are updated. For this reason, we no longer disinfect our mail, but we continue to recommend masks. Nobody can ever be 100 percent sure that these recommendations are correct – new knowledge may emerge – but they reflect the best information we have.

The same is true of the smoking paradox: before the pandemic, the evidence was that smoking was doing your lungs no good. With new – good – information, the probabilities could have been updated and shifted towards the extraordinary claim that smoking is protective.

And that’s the second point: Was that any good evidence at all?

It was not.

First, most of the publications on the smoking paradox had not been peer-reviewed by other scientists when they were reported. While a good number of publications were peer-reviewed, others were withdrawn after it became clear that they were funded by the tobacco industry. Posting before posting is a great way to get information ready quickly. it is not a good thing to make sure the information is solid.

Second, most of these studies were small. While this is not a death bell, it does mean that the evidence should be treated with caution. In other words, probabilities can update, just not by much.

This makes sense intuitively: if you get 999 heads on 1,000 coin tosses, you are pretty sure that the coin has been tampered with. If you’ve got two heads on three flips, you’re a lot less sure. The studies pointing to the smoking paradox had sample sizes ranging from teenagers to hundreds. The British study that disproved this had 421,000.

Finally, and subtly, the smoker’s paradoxical studies asked a different question than it should have. They asked: “How many of the people currently in hospital smoke?” It is different with: “How likely is it that smokers in the population are hospitalized compared to non-smokers?”

The first question relates to people who have already been admitted and survived long enough to be studied. In other words, just like Bertand’s coin boxes, approval has already been granted and there are many reasons why smokers were not included in this group. Perhaps they died faster than non-smokers, so they couldn’t be counted. Perhaps they were released to the hospice at another point in time. The British study, on the other hand, looked at the entire population and removed this bias.

So I would argue that science has not misunderstood the smoking paradox. It was an interesting finding that resulted in widespread extraordinary harm. And if Covid doesn’t teach us otherwise, it should teach us to make extraordinary demands – about smoking, vitamin D, zinc, bleach, iodine gargling or the atomization of hydrogen peroxide – to high standards.

Science moves slowly. Extraordinary claims are not. To paraphrase Jonathan Swift, they fly with them while evidence lags behind them.

Mark Shrime is Chair of Global Surgery at RCSI University of Medicine and Health Sciences. A version of this article first appeared in The Conversation and was reprinted by special arrangement.