Ron DeSantis-cited article to oppose mask-wearing for kids falls apart under scrutiny

As Florida has sunk ever-deeper into its third major wave of Covid-19, Republican Gov. Ron DeSantis has dug in his heels on his relatively hands-off approach to the virus, which many say has allowed it to spiral out of control.

Hospitalizations in the state have reached a recent peak of around 17,000, straining local resources. Average daily deaths from the virus recently approached 250, significantly above the worst days of the previous waves. With the spread uncontrolled, many are nervously awaiting the end of the summer and the start of the school year, when millions of children will attend schools packed with classmates, all breathing the same air.

But DeSantis has opposed one of the most seemingly common-sense measures to help reduce the spread of the virus among kids: school mask mandates. He's gone so far as to threaten school district funding if they disobey him, and he doesn't seem deterred by the mass outbreaks and quarantines already affecting thousands of Florida students. (On Friday, a judge ruled against the DeSantis policy.)

When asked about his policy this week, he lashed out at the press for its advocacy of masks for kids — which also happens to be what the Centers for Disease Control and Prevention recommends. To defend his position, he cited a recently published article.

"The evidence doesn't support it," he said. "There's a great article, actually in New York Magazine of all places, that went through this 3,000 words, very thorough, view of all the evidence and lack of evidence."

I was intrigued by DeSantis's citation, having not seen the article. So I decided to track it down and see what kind of case it made. Unfortunately, for DeSantis, it's a compellingly written but ultimately unpersuasive argument — one he was surely too willing to accept rather than actually scrutinize.

The headline is modest enough, claiming simply that "The Science of Masking Kids is Uncertain." Indeed, much of the science around the virus remains uncertain. But the piece, written by writer David Zweig, is really a polemic against having kids wearing masks, strongly suggesting that there's no evidence to support their use and that there's significant risk of harm from using them.

He wrote vaguely about Europeans being less inclined have students wear masks, and said "there's no evidence of more outbreaks in schools in those countries relative to schools in the U.S., where the solid majority of kids wore masks for an entire academic year and will continue to do so for the foreseeable future." But he doesn't actually provide any figures or detailed comparison of these cases, so this really isn't much to go on. He also pointed out that the World Health Organization is much less enthused about students wearing masks than the CDC is — though he neglected to mention that WHO was significantly behind the curve on recognizing the importance of masks generally for reducing the spread of the coronavirus.

The bulk of Zweig's argument actually relies on evidence from the CDC. He argued that one of its own studies of COVID-19 spread in Georgia elementary schools last fall, which he called "both ambitious and groundbreaking," actually shows that mask mandates did not "have a statistically significant benefit." He adds: "In other words, these measures could not be said to be effective."

He essentially accused the CDC of covering up its own finding, though he doesn't explain why it would do that. In fact, it's Zweig who is hiding the ball.

You wouldn't know it from reading his article, but the study Zweig cites actually found that the schools it studied with mask mandates for students had 21 percent less spread of the virus than in schools where masking was optional. This amount was comparable to the study's primary findings, which was that COVID-19 incidence was 37 percent lower in schools requiring masking for teachers and 39 percent lower in schools that improved ventilation.

It is true, as Zweig wrote, that the study's authors concluded the finding about student mask mandates wasn't statistically significant. But he slyly suggests this was a finding that the masks don't make much of a difference for kids. That's not so. It's completely possible mask mandates reduced the incidence of Covid by around 20 percent — a significant effect, to be sure — but the study itself wasn't able to make this causal claim.

Here's how the study itself actually addresses the finding:

The 21% lower incidence in schools that required mask use among students was not statistically significant compared with schools where mask use was optional. This finding might be attributed to higher effectiveness of masks among adults, who are at higher risk for SARS-CoV-2 infection but might also result from differences in mask-wearing behavior among students in schools with optional requirements. Mask use requirements were limited in this sample; 65.1% of schools required teacher and staff member mask use and approximately one half (51.5%) required student mask use. Because universal and correct use of masks can reduce SARS-CoV-2 transmission (6) and is a relatively low-cost and easily implemented strategy, findings in this report suggest universal and correct mask use is an important COVID-19 prevention strategy in schools as part of a multicomponent approach.

It's understandable that some people confuse "not statistically significant" effects with "virtually nonexistent" effects. But it's not helpful for a writer making Zweig's case to ignore the magnitude of the finding.

"One of the biggest problems with interpreting findings that are not 'statistically significant' is distinguishing between the lack of a true association versus missing a true association due to inadequate power or due to other sources of epidemiologic bias," Brandon Guthrie, a professor of epidemiology and global health at the University of Washington, wrote to me in an email. "A 21% reduction in risk, if it were true, would be meaningful, and therefore I would conclude that the study was inadequately powered to answer this question."

We should also be careful to distinguish between the effects of students wearing masks and the the effects of students being mandated to wear masks. It could be possible that schools without masking mandates nevertheless have a high rate of mask-wearing. It might then be hard to distinguish the effect of the mask mandate in a study, but it would certainly be wrong to conclude from this that the masks themselves were having no positive effect.

Zweig emphasizes that he looked for other comparable studies, directly comparing schools without mask mandates to those without, but he came up empty. He thinks this is telling, but he's really asking for too much.

"The reality is that we don't have randomized trial evidence that compares mask usage for prevention of the COVID-19 virus to non-usage, holding everything else constant, in real-world settings at a population level," Guthrie explained to me. "Such a study would likely be impossible to conduct ethically, [and] even if it were ethical, would be nearly impossible to conduct practically. In the absence of that evidence, we have to rely on interpreting multiple lines of indirect evidence."

There's plenty of evidence that masks reduce the spread of Covid-19 — indeed, the study at the center of Zweig's argument offers strong support for the effect of masking adults at schools. If masks reduce the spread of the virus among adults, why shouldn't we expect that they'll do the same for kids?

Zweig does have a response to this argument: "While masks offer some protection for adults in many environments, as the adage in pediatrics goes, children are not little adults. Medicine is littered with examples of adult interventions that don't translate to children."

This point is worth considering, but it's less compelling than Zweig suggests. Masks reduce the spread of Covid because it's a respiratory illness that transmits via exhaled droplets. Masks are barriers inhibiting that process. It's hard to give a principled reason to think a simple method of interfering in the basic avenue of transmission would be effective for adults but not for children. It's not as if Covid infects children through the ear canal rather than the airway.

Zweig indicates that the physiological differences between children and adults may explain why the CDC study of Georgia schools didn't find a statistically significant benefit of masking students, though it did find one for adults. But as I noted, the study did find a large association between student mask-wearing and reduced viral spread, even if a causal inference couldn't be made. The fact that masking was shown to reduce spread from adults should actually heighten our credence that the reduction in spread for students under a mask mandate was a result of the intervention. Indeed, that's essentially the argument the study's authors made.

It is true that studies seem to show that children spread Covid less readily than adults do. They also get sick less and have a much lower fatality rate. And masks may not work as well on children for behavioral reasons — for example, they may take them off or adjust them more frequently, or they may not be comfortable wearing tighter-fitting masks that better reduce transmission. These may make the impact of mask mandates for kids less impactful — but that's no reason to think the effect is negligible.

Even while Zweig discounts the evidence in favor of masks, he exaggerates the evidence on the downsides. At one point, hewrote:

"There are very good reasons that the World Health Organization has repeatedly affirmed their guidance for children under 6 to not wear masks," said a pediatrician who has both state and national leadership roles in the AAP but who wished to remain anonymous because they did not want to jeopardize their roles in the organization. "Reading faces is critical for social emotional learning. And all children are actively learning language the first five years of life, for which seeing faces is foundational," the pediatrician said.

But while seeing faces in the first five years of life may indeed be quite important, that's not really relevant to whether the individual kid is masked — what matters is whether they're seeing masked faces. And most of the first five years of a child's life aren't in traditional k-12 schools. Thankfully, in much of their time at home, in outdoor settings, and in smaller groups, quite young kids should have plenty of opportunities to interact with people with uncovered faces, even during this pandemic. And it's definitely good for teachers to find multiple opportunities for students to interact maskless throughout the day outdoors, where the risk from the virus is particularly low. And teachers should strive to accomodate any students who may, for whatever reason, have a particularly difficult time wearing masks.

Everyone should agree that we want the coming school year to proceed as smoothly as possible and with as few virus-related interruptions as possible. That means using every tool we have to reduce outbreaks of Covid, including vaccination for those who are eligble, increased ventiliation when possible, and, indeed, masking in large groups whenever it's feasible.

And of course, a big factor in any discussions about mitigating Covid spread in the fall is the Delta variant. Most of the studies we've seen on Covid examined previous variants of the virus. And the major difference with the Delta variant is that it spreads even more easily than the initial form of the virus.

Zwieg dismisses concerns about Delta, writing:

A common argument right now is that the emergence of the Delta variant changes everything. Currently, some regions of the U.S. are seeing a surge of infections and hospitalizations among young people. But the numbers coming out of Britain continue to suggest that Delta is not more virulent — that is, it does not cause more severe illness on an individual basis to unvaccinated people — despite being more contagious. A pediatric immunologist at a major university hospital who was not authorized to speak publicly said, "It is not biologically plausible that the same variant somehow is more dangerous for kids in the U.S. than it is in the U.K."

More broadly, Schecter-Perkins said, "I don't think that Delta changes the calculus because it still seems clear that it doesn't cause more severe disease, so it still doesn't change the fundamental question of 'What are we trying to achieve by masking kids when they are still extremely unlikely to suffer from severe illness or death if infected?' And the adults in their lives have the opportunity to be vaccinated and also protected so we don't need to worry about transmission." The pediatric immunologist said, "Even with a new variant, the onus is on those who recommend masking kids to robustly demonstrate a meaningful benefit, especially when the pre-Delta study of the Georgia schools did not find one, and when there are obvious socio-emotional and educational harms from masking children for this unprecedented duration of time."

But increased trasmissibility is precisely the concern about Delta, and it could have a major impact on viral spread at school. More kids getting Covid will certainly mean more kids getting severely ill and dying. Those numbers may, nevertheless, remain relatively quite small. But the virus's increased transmissibility may turn unmasked schoolchildren into signficant vectors of viral spread in the community. And you can't dismiss this concern because the adults in the community have an option to be vaccinated. Uncontrolled spread is devastating for a community, for the vaccinated, the unvaccinated, and immunologically compromised. Every part of the community needs to be a part of the effort to keep the virus under control.

Further, Zweig cites one of his sources arguing once again that there's "obvious socio-emotional and educational harms from maksing children," but he doesn't provide any signficant evidence for this claim — certainly none that meets the standard of evidence he requires for claims about school mask mandates.

Meanwhile, many experts argue that the new forms of the virus mean we should be as vigilant as ever.

"The presence of the more transmissible [Delta] variant is a strong reason to emphasize basic strategies like universal mask usage in schools, as well as in other public indoor spaces," Guthrie told me. "Masks, while not always the most enjoyable experience, are not particularly inconvenient in most settings, and when combined with vaccination and other minimally disruptive measures, they are an important part of making classrooms safe."

A recent CDC study highlights the danger. It found that an unvaccinated California teacher brought the virus into her classroom while symptomatic, and despite masking rules, she read aloud to the students with her face uncovered. The kids were supposed to be masks, and the students reportedly said they complied with this rule at a high rate. Nevertheless, fifty percent of the students were infected, and some of them brought the virus home to their parents and siblings. A nearby classroom of 18 students also had six cases; one of the students had recently had a sleepover with two other students, and all three ended up with Covid. There were also four other student cases detected in other grades, each of whom had siblings in the the class with the initial outbreak. No other infections were reported in those classrooms, potentially because the student masking prevented further spread.

Of course, the biggest lesson here is that teachers and anyone else who is eligible should be vaccinated, and people with symptoms of Covid should self-isolate. But it's also clear from this incident that the virus can spread quickly within schools, and it can make it's way from there out into the wider community. The students in the classroom were reportedly masked, distanced, had open windows and doors for ventiliation, and had an air filter at the front of the room. Without these strategies, the outbreak might have been worse. And it's hard to be confident that any one of the strategies is making a huge difference on its own or is going to be a silver bullet — but we shouldn't be quick to abandon any strategy that we think has a good chance of providing substantial protection.

Indeed, all this reflects the lesson of one other study Zweig cites in Science Magazine, based on survey data, which he did admit is suggestive of a positive impact from mask mandates.

"We found that when seven or more mitigation measures were in place, the positive association between in-person schooling and COVID-19 outcomes disappeared," it said. "Among those reporting seven or more mitigation measures, >80% reported student and teacher mask mandates, restricted entry, extra space between desks, and no supply sharing, and >50% reported student cohorting, reduced class size, and daily symptom screening."

On the other hand, we shouldn't be doctrinaire about any of this. It's possible that on the whole, Delta will spread more easily among kids than previous variants, making masks even more important in these settings. But it's also possible that precisely because Delta spreads so easily, it doesn't actually do much to reduce the spread when students are mixing among themselves for many hours a day. We should pay close attention to what the emerging evidence shows and let it guide our actions moving forward. The Science Magazine study, for example, found that desk shields are unlikely to be helpful in schools and may actually increase the spread of the virus. They may be worth abandoning altogether.

But the fact is that public health officials are trying hard to evaluate the evidence. Contrary to Zweig's implications, there's little indication they are hiding their findings. They're making tricky judgments in an uncertain environment when we may not always be able to get th exact kind of evidence that we would most like to see. And for now, there's ample evidence to ask the kids going to school this fall to mask up when the risk of viral spread is significant, for themselves and for the good of the community.