Is it possible to reopen school buildings in the fall in a way that keeps kids, educators, staff and their families and communities safe from Covid-19? Is it possible not to do so without harming them in other ways? Already, school closures have set children behind academically. More than 20 million children rely on school breakfasts and lunches. Too many parents face the choice between losing their jobs or leaving their children at home unsupervised. Vaccination rates for various dangerous diseases, typically required before students can attend school, have plummeted. Isolating children from their peers exacts social and emotional costs, which differ by age group and are nearly impossible to quantify. And whether schools reopen or remain closed, the risks are borne disproportionately by low-income communities and people of color. “This is really one of the most perplexing and complex issues I’ve ever faced in 40 years,” says Dan M. Cooper, a professor of pediatrics at the University of California, Irvine.
A flood of guidance has been issued in recent weeks, much of it urging schools to reopen and suggesting safety precautions. Media outlets as well have relayed reams of often conflicting expert advice on how to weigh risks and benefits, to individuals and to society. In every case, that calculation is constrained by major gaps in our understanding of how Covid affects children and those in contact with them. Strong evidence suggests that children are much less likely than adults to get sick or die from the virus. (By July 9, data from most of the U.S. showed that nearly 242,000 children had tested positive for Covid, representing 8 percent of cases, the American Academy of Pediatrics reports; they account for fewer than 3 percent of hospitalizations and fewer than 1 percent of deaths.)
But are children less likely to be infected, or just less likely to show symptoms? Does the virus behave differently in grade-schoolers than in high-schoolers? What factors determine whether children become seriously ill? And, perhaps most crucial for schools, what are the odds that students will transmit the virus to one another or to adults?
One reason scientists have a lesser understanding of how the coronavirus acts on children is that in March, at the outset of the pandemic, most stricken countries simultaneously closed schools, shuttered businesses and urged people to stay home, making it nearly impossible to separate out the effect of school closures on rates of transmission in the community. During a stay-at-home advisory in Switzerland, researchers from Geneva University Hospitals tried to determine how vulnerable various age groups were to infection. Beginning in April, they adapted a health study already underway to test residents for coronavirus antibodies. Subjects came in weekly and were invited to bring everyone they lived with who was at least 5 to be tested, too. The results from more than 2,700 participants over five weeks, published in The Lancet in June, showed that children ages 5 to 9 and adults over 65 were significantly less likely to test positive than those between the ages of 10 and 64. Of the 123 children in that age group, 21 were exposed to an infected household member, but only one developed antibodies.
Large-scale randomized testing and contact tracing over time, which would give a more complete picture of who transmits the virus and how, hasn’t been done yet in schools. In July, in the journal Emerging Infectious Diseases, researchers from the Korea Centers for Disease Control and Prevention published the results of tracing more than 59,000 contacts of 5,706 coronavirus patients. Children younger than 10 were found to have transmitted the virus much less than did those between 10 and 19, whose transmission rate was equivalent to that of adults. But only 3 percent of patients in that initial cohort were 19 and younger, and their having been tested probably means they presented symptoms. It’s still unclear how asymptomatic children, who are hard to identify, might spread the virus; it’s also unclear if there are differences in transmission between the ages of 10 and 19.
“A lot of the data we’re getting from different sources is messy and not necessarily pointing in the same direction,” says Nicholas Davies, an epidemiologist at the London School of Hygiene & Tropical Medicine. He and colleagues used a statistical method called Bayesian inference to test several hypotheses about children and Covid. These included the premises that kids are being infected but not showing symptoms and that kids are less susceptible to infection. Based on epidemiological data from China, Italy, Japan, Singapore, Canada and South Korea, the researchers concluded that both premises were probably true, to an extent. Their findings, published in Nature Medicine in June, estimate that people under 20 are about half as likely as older age groups to become infected, and that among 10- to 19-year-olds who do get the virus, only 21 percent will have clinical symptoms. They couldn’t make finer age distinctions, nor say how likely any children are to infect others.
None of these studies directly addresses the impact of reopening schools on the spread of Covid. In fact, when researchers from the University of Washington departments of global health and epidemiology began compiling a summary of models from 15 other countries where students have returned, they found “very few” scientific publications on the topic and relied primarily on news reports. In nearly all countries, they observed, schools adopted safety measures, including face masks and social distancing. None of the countries (except Sweden, which kept many schools open) resumed classes before national rates of infection had significantly declined; there is no evidence to say what the outcome of opening schools would be in areas of the U.S. where the virus is surging. In Germany, where infection rates were higher than in other European countries, the return of older students accompanied an increase in infections among one another but not staff, according to a preprint led by researchers at the University of Manchester and Public Health England.
In Israel, students and staff wore masks after schools reopened in early May. But several weeks later, those rules were relaxed. According to Haaretz, outbreaks began soon after, exposing thousands at schools to infection, causing many of them to close down again. There’s “not clear cause and effect” between the removal of masks and the outbreaks, says the summary’s lead author, Brandon Guthrie, but it’s “circumstantial evidence” that they offer some protection in classrooms. It also reveals how unenforceable the health guidance schools are receiving can be.
Cooper, co-author of a commentary in The Journal of Pediatrics in May that highlights the need for collaboration between local schools and public health officials, believes, in general, that “schools need to reopen, and we need to study what happens in the schools very, very carefully.” The C.D.C. could be “quite helpful,” according to Anita Cicero, deputy director at the Johns Hopkins Center for Health Security, if it “put out a model protocol” for researchers trying to track Covid cases that emerge in schools “so everyone is collecting data the same way.”
In May, the N.I.H. initiated a study to test thousands of children and their families over six months to see who gets the virus, whether it’s transmitted within the household and who develops Covid, while collecting information about participants’ recent activities. That’s the kind of detailed data collection needed to help determine under what conditions schools are likely to endure outbreaks or contribute to community spread. But none of that data will help us in time for the start of the school year. Instead, without the ability to consistently test students, get quick results and trace contacts, it will be impossible for schools to tell who has the virus and whether it’s circulating on campus; when students and staff inevitably get sick, individual schools will have to debate shutting down or staying open without any more useful information to guide them than they have now. To all of America’s failures in the Covid-19 crisis, we should surely add this one: the inability to get schools the tools and data they need to strike the best possible balance between education and health.