Professor Carl Heneghan:

Again, this is about the evidence. In 2010, at the height of the last pandemic, there were six published trials of about 4,000 people. Since then, we have not addressed the lack of evidence and closed that graph. In the intervening ten years, there have been about another six trials. If one looks at the 12 trials together, what they show in healthcare professionals is that masks, gloves and a combination of PPE reduces the risk of infection.

When one goes into the wider population, there is a small bit of evidence that shows that if one has influenza in one’s household - a child, for example - and if one wears a mask for one week in the house, one can reduce one’s risk of influenza or likely illness by about 10%. However, one has to completely adhere to mask wearing for the whole week. If one stops adhering to it, as 50% of people did, one loses all the effect. That is one of the problems.

The second issue is that the evidence comparing cloth masks to surgical masks or the N95s shows clearly that cloth masks are worse and may actually increase the risk of infection. Therefore, that is why they are not recommended in hospitals or in health professional settings.

What happens in these situations of uncertainty is that the opinion divides. Someone thinks mask are a good idea while someone else does not think they are and that people should not wear them. That is why we end up with people proposing them more and more. They say people should put them on in schools, pubs and shops. However, there is no clear evidence. They use observational data to inform their decision. If one looks at what has happened in the UK, for instance, it put masks in on 24 July. They were supposed to reduce the risk of infection by 40% over the next two weeks. In fact, infections detected have gone up. In effect, people are not looking at the evidence.

When Norway looked at this, it said that at low circulation, the public health consequences were so minimal that it was not clear they worked and, even if they did work, it reckoned about 200,000 people would have to wear a mask fully for a week to prevent one infection. That is how one has to think about the uncertainty. If one is going to put a policy in place, that is fine. What one cannot do, however, is say it is evidence-based because when people talk about the evidence, they have again cherry-picked low-quality, observational evidence to suit the argument. However, they have not picked further observational evidence. For instance, if one puts masks in, what is one expecting to happen to the case definitions and the reductions in the next two to four weeks to show it was a worthwhile policy to enact?