© Provided by Real Clear Science

We have heard so many interpretations of research on the value of masks, that another may be unwelcome. A comprehensive review of the evidence from the prestigious UK based Cochrane Collaborative concluded we don’t know the benefits of masks because the right studies have not been done. Yet, this straightforward assessment has been obscured by claims the science proves masks do not work and counterclaims that mask use is supported by the best science. How can the same science be so contradictory?

Let’s look at what the report said. The review considered only clinical trials that randomly assign volunteers or communities to a treatment or a control group. In most of the studies, “treatment”, was being asked to wear a mask. In cluster randomized trials, an entire community may be advised to wear masks while a control community is not so advised. Averaged across 12 trials that were conducted during the COVID pandemic, wearing a mask made little or no difference in acquiring respiratory illness.

With these results, scientists, physicians, and journalists were off to the races. The headline of a New York Times editorial read “The Mask Mandates Did Nothing.” One distinguished physician proclaimed that mandates had been enacted without any scientific justification. A few weeks later, another Times commentary boldly asserted “the science is clear” – masks work.

As scientists, we looked at the question the study was designed to answer. We were surprised to learn that none of studies directly tested the effects of mask mandates. In these trials, people were asked to wear masks. There was no mandate. For example, one study in Bangladesh involved more than 300,000 people. In some randomly chosen villages, public announcements encouraged mask use while other villages were used for comparison. But, only 40% of those in villages where masks were promoted actually used them (and 10% in the control villages wore masks). Being advised to wear a mask did not significantly reduce COVID infections. Does this mean that masks didn’t work? No. It means that advising people to wear masks didn’t work. (By the way, among older people there actually was a 35% reduction in COVID infections.)

The problem in the Bangladesh study was not unique. In most of the studies, many people did not comply with the request to wear a mask. In an influential Danish study, only 46% of “mask users” actually wore the mask as recommended. In another well-known study, those assigned to wear masks did so only an average of 2.5 days per week and for less than 4 hours per day.

Few of the studies in the review were designed to answer the question we usually confront: Do masks reduce COVID infections among people who are highly motivated and willing to accept masking advice? What advice does the science provide for this specific question? Unfortunately, not much. In public health research, the term fidelity describes how well people accept and use the treatment to which they were assigned. For the masking studies, fidelity was consistently dreadful. Unfortunately, this introduces a serious bias. People who adhere to treatment almost always achieve better health outcomes for reasons that may have nothing to do with the intervention. They may be engaged in lots of other healthy behaviors that lower their risks. It may look like they benefited from the treatment, but their health may be better explained by other things they do to stay healthy.

To confront this bias, statisticians insist on an “intention to treat” strategy that counts someone as a member of the mask group based on where the randomization sent them — whether or not they actually used the mask. The method does reduce bias. But, when more than half the people asked to use a mask ignored the request, studies fail to address the core question: Will my decision to consistently wear a mask prevent me from getting COVID?

So where does this leave us? Years of study and, more recently, weeks of commentary, fail to inform us about the value of mask wearing. The crucial question of whether you should wear a mask in the grocery store remains unanswered and surprisingly understudied. For some people, ditching masks is justified by the absence of compelling evidence that masks prevent infections. But nearly all the commentaries ignored the Cochrane conclusion—current evidence does not answer most of the key questions. The conclusion that masks don’t work should more accurately be stated as “studies have not yet demonstrated that masks work”. And, the fact that there are serious flaws in studies that fail to show masks work does not support the conclusion that masks do work after all. Even though both sides cite science as their ally, current evidence does not tell us with confidence whether masks work or are a useless bother. For that we await better research. But, this should not be viewed as a research failure. Acknowledging what we don’t know is good science practice. It directs the scientific agenda and informs legitimate public discussion.