This article was first published in The Montreal Gazette.
A new analysis published by Quebec’s public health institute has reviewed the data on the effectiveness of masks in preventing COVID-19. They concluded that masks offer up moderate protection in preventing viral spread in home and hospital environments, but said they found insufficient evidence about their effectiveness in the broader community setting. Unfortunately, some people are going to interpret these findings to mean that “masks do not work.” Such an interpretation is very wrong.
It’s important to understand what exactly was done in this evidence review. While the authors did not produce any new data, they scanned the literature to see what had already been published, assessed the quality of the existing data and then summarized it mathematically in what is called a meta-analysis.
Most of the existing literature on mask use relates to the flu or SARS or other viruses rather than to COVID-19, given that COVID-19 is a new virus that has only been present in North America for about six months.Another caveat is that most research about viral spread is generally done in hospital or health-care settings. Organizing research projects in hospitals are easier than doing so in outside settings and recruiting research subjects among medical personnel is easier than recruiting them from the general population. The pivotal Doll and Hill study that first established the link between smoking and lung cancer was a survey conducted among British doctors (back when many doctors still smoked).
A lot of the existing research was also done in home settings where researchers could track whether sick patients infected household contacts. Most parents know that whenever their children catch a cold, they ultimately will catch it too, because viruses spread easily within homes, where you have repeated and prolonged close contact with the same individuals. All this to say, there is a lot of data on mask use, but not specifically on mask use in the wider community setting. Purists would say that until this specific question in this specific context is tested and studied, we must be deliberately agnostic about the benefits of masks for the general public. I myself firmly believe that evidence must guide our clinical decisions and we must be certain that we have tested any medical intervention before recommending it to the public. But we must also be careful not to let the perfect become the enemy of the good. An infamous tongue-in-cheek paper in the BMJ in 2003 reviewed the literature and found no randomized trials to support the use of parachutes by skydivers. The point of the paper was not to recommend unencumbered free-fall among parachutists, but to remind the medical community that “under exceptional circumstances, common sense might be applied when considering the potential risks and benefits of interventions.” These are indeed exceptional times.
We cannot toss people out of planes without parachutes to prove a point, just as we cannot take two groups of people (one with masks and one without) and deliberately expose them to a deadly virus to see how many in each group survive. In some situations, especially in a new global pandemic, definitive research on certain topics is logistically and ethically impossible. It is critically important that we test any medication or vaccine to be certain it is both safe and effective before administering it to patients.
But masks have no major downside apart from the mild inconvenience of wearing it and having to wash it afterward (if it is a cloth mask). I have no randomized trials showing that parachutes prevent broken bones in skydivers, and I have no randomized trials showing that masks prevent viral transmission in the general population. But I’m pretty sure that both of those things are true.