Since the beginning of the pandemic, one of the most contentious questions has been under what conditions it might be safe to reopen schools. Families are distressed about children suffering academically and emotionally at home, and leaders are anxious to get parents back to work, partly to aid economic recovery. The absence of good information about how the coronavirus spreads, along with a lack of federally led testing and contact tracing, has fueled massive amounts of distrust, with some suspicious that reopened schools are less safe than leadership purports and others convinced closed schools are safer than leaders claim. Teachers and communities have organized protests, both to close schools and to open them.
Now, higher-quality research is beginning to emerge, shedding light on the relationship between reopened schools and the spread of the virus earlier this fall. Two new studies in the U.S. that were published in the last two weeks suggest that when community transmission is low, reopening school buildings, at least when schools are not operating at full capacity, does not contribute much to the virus’s spread. But the risks change, the researchers found, when community transmission is higher. A third study, published in early December, found that reopening Florida schools led to increased infections among school-age children, particularly among high schoolers. Florida is one of the only states to provide incidence rates by age and county.
One of the studies, published Monday by two economists and one epidemiologist at Tulane University, looked at the effects of school reopening on Covid-19 hospitalization rates. The authors believe it’s the first study worldwide to explore this relationship, as nearly all prior research has compared school reopenings to case positivity rates — a less reliable but more accessible measure. While hospitalization rates don’t capture those who may still be suffering from long-term symptoms of Covid-19 or who got very ill but never sought hospital treatment (perhaps due to lack of health insurance), the researchers say it’s still a more robust way of measuring real sickness in a community than case rates, which include those who are asymptomatic or have very mild illnesses.
The study found that in counties where new Covid-19 hospitalization rates were below 36-44 per 100,000 per week — which was roughly 75 percent of U.S. counties this past summer — in-person school reopening did not lead to an increase in hospitalizations. Researchers found, however, that in places where new hospitalizations exceeded that threshold, the school reopening picture was far less clear. And because even slight increases in transmission can have significant effects, more caution, they said, is warranted.
The study comes at a moment of great upheaval, as the virus reaches its most out-of-control point in the U.S., and a new, more transmissible variant, first detected by British scientists about a month ago, is spreading across the globe. On Monday in England, just a day after reopening schools and insisting they were safe to attend, Prime Minister Boris Johnson announced a new strict lockdown to contain the virus, including closing all schools for at least the next six weeks. Scotland made a similar announcement earlier that day, with its First Minister Nicola Sturgeon extending remote learning for another month, citing the “simply too high” levels of community transmission and the “significant uncertainty” about how the new variant impacts young people. (In both countries, in-person learning will remain for vulnerable students and children of essential workers.)
While the Tulane study only looks at rates between late August and October, a period when the researchers say they could get the cleanest estimate, the researchers noted that hospitalization rates have spiked much more over the last month amid the new surge. Study co-author Douglas Harris told The Intercept that he and his colleagues recently did a new calculation based on the latest available hospitalization rates, from the week of December 14-20. They found that 58 percent of counties would be safe to reopen under their threshold benchmark, down from 75 percent over the summer.
By controlling for variables like college reopenings and state policies around social distancing, the researchers felt confident attributing the rise in hospitalization rates to K-12 schools, as opposed to other potential factors. They relied on recently released hospital data from the U.S. Department of Health and Human Services and health insurance claims data provided by a private organization, Change Healthcare.
A second study, published in late December, looked at the impacts of Washington state and Michigan school reopenings on Covid-19 transmission through November. The researchers focused on positivity rates, and like the Tulane study, they found that in areas where transmission rates were low, in-person schooling didn’t appear to contribute to the virus’s spread. But when community rates were higher, reopening schools was linked to increased spread. Michigan and Washington have been collecting data on each school district and making it publicly available, something many states have not done.
In Michigan, the threshold linked to increased risks from reopened schools began at 21 new daily cases per 100,000, and in Washington, it was observed at just five new daily cases per 100,000. Today nearly every county in both states exceeds those levels, and the majority of the country is also well beyond those November rates.
Katharine Strunk, an education policy professor at Michigan State University and one of the study co-authors, told The Intercept they could not determine why the thresholds varied between the two states, though they assume differences across local contexts play a role. As a result, she and her co-authors avoided putting forth one national positivity threshold for school reopening risk, unlike what the Tulane study did with hospitalization rates. “We looked for explanations, and we don’t have them,” Strunk said. “And that’s part of why we don’t want to give one number.”
The third study, focused on Florida, found that in counties among schools that reopened, student infections rose, particularly among high schoolers aged 14-17. (The research could not distinguish within counties between students who returned to school and those who opted for remote learning.) The analysis was conducted by researchers at Ben-Gurion University of the Negev, Harvard Medical School, and Tel Aviv University and published in the medical journal Archives of Disease in Childhood.
— Rachel Cohen (@rmc031) January 6, 2021
Some experts insist that schools can be safe to reopen even amid very high levels of community transmission if infection control strategies like universal masking, asymptomatic testing, and robust contact tracing are in place. These leaders argue that if the right investments are made into school safety, then in-person learning will only have “limited impact” on community transmission rates, and the emergence of the new B117 variant need not derail reopening plans. Others are more vocally acknowledging that yes, reopening schools now will likely accelerate the spread of the virus in the community, but they maintain that the social, emotional, physical, and academic benefits to children could still outweigh the transmission risks, particularly for disadvantaged children. Evidence is mounting of students falling behind in their classes.
On Monday, an op-ed in the Washington Post illustrated how the terms of the debate have shifted. Two epidemiologists from Johns Hopkins University wrote that “it strains credulity to claim that schools play no role in driving viral spread” and noted that multiple studies have shown school closures were associated with “substantial reductions in community spread.” Rather than insisting schools are safe, they argued that the benefits of in-person learning were so important that leaders should be “honest” about the risks posed by Covid-19 in schools so communities could more effectively tackle them. The authors recognize this is easier said than done. “Frank and open conversations among stakeholders … might be even more difficult than finding time and money for mitigation efforts,” they wrote, identifying the charged politics around the issue.
Investments in school mitigation measures like ventilation upgrades and routine testing would help, and the Biden administration, which wants to open most elementary and middle schools by the end of its first 100 days, is seeking tens of billions of dollars to aid that goal. California Gov. Gavin Newsom also recently announced a plan to incentivize school re-openings by offering districts a base rate of $450 more per student, but given the state of the virus in California, it’s not yet clear whether districts will bite. Vaccinating teachers, though, could hasten the process and reduce opposition from educators and their unions. In some states, like Tennessee, Oregon, and Colorado, school employees have already been granted priority vaccine status.
Harris says he thinks his study can help inform policymakers in the spring, when the new surge of the virus has hopefully dissipated. And while the spread of the new variant might be much greater by then, Harris thinks a strain that spreads more easily will likely just mean lowering their hospitalization threshold of 36-44 per 100,000 somewhat.
Neither the Tulane study nor the study on Washington and Michigan was able to glean which reopened school districts used which school mitigation strategies. Experts have been urging universal masking, though debate exists over whether six feet or three feet is needed for student social distancing. Quality building ventilation and routine in-school testing are also highly variable across districts and schools.
“Right now, there’s not a lot of data across or within states on that, and that is a real shame,” said Strunk. “There’s definitely no federal push to collect it.”