For the most part, school reopenings in the fall did not appear to contribute to increased hospitalization rates due to COVID-19,
The finding adds to a growing body of evidence suggesting that schools did not play much of a role in fueling infections when community transmission rates remained relatively low. It is the first study to use hospitalization as its key health measure, a research advance that avoids some of the problems with using test-positivity counts as a proxy for COVID-19 spread.
But in places where community spread was higher, the researchers found that the link between schooling and health effects grew murkier, with no clear pattern in the results, a red flag of sorts as schools consider expanding in-person learning options in the midst of a third surge of record-breaking rates of COVID-19 from coast to coast.
鈥淕iven the recent spike in hospitalizations in recent months, policymakers should be cautious. It may not be safe to reopen schools where the virus is already widespread,鈥� said Engy Ziedan, an assistant professor of economics at Tulane who studies health care and a co-author on the study, in a statement.
Nor does it address what could happen if a new, takes root across the United States.
In all, the findings highlight what has been one of the most frustrating aspects for school district leaders desperate for objective guidance: The lag in the knowledge base on the virus in schools.
The pandemic continues to evolve rapidly, while research on it has necessarily followed several months behind. What made sense months, weeks, or even days ago might not make sense as rates continue to change, and as beleaguered school officials face competing demands from community members, parents, and staff.
Does Reopening School Increase Hospitalization Rates?
The new study, released by the National Center for Research on Education Access and Choice at Tulane University, comes right as most schools resume class from their winter holiday breaks.
For school reopenings, the researchers looked at data on schools from Education Week鈥檚 fall 2020 school reopening sample of some 900 districts, plus larger samples based on two commercially available lists. They combined that data with health outcomes based on health insurance claims data from Change Healthcare, a health-care technology company, and hospital data from the U.S. Department of Health and Human Services.
That is an important breakthrough in the research because of problems associated with the commonly used yardstick of test-positivity rates, which are highly dependent on county testing systems鈥� capacity and contact-tracing protocols.
Hospitalization for COVID-19-related diagnoses, on the other hand, is a direct measure of illness related to the pandemic.
The researchers used statistical methods to examine the link between hospitalizations and the mode that the school district reopened in鈥攈ybrid, in-person, or remote鈥攚hile controlling for factors like state mandates on masks and social distancing. They tracked counties鈥� health data week by week, looking at reopening decisions made in mid-August through October. They also developed a new measure of collective bargaining strength to control for how teachers鈥� unions shaped some of the reopening decisions. (Prior research had used district size, an imperfect proxy.)
In all, the research found that in-person schooling had no effect on hospitalization rates in counties in which the number of new COVID-19 hospitalizations, pre-opening, fell below 36 to 44 per 100,000 people per week. About 75 percent of counties were at or below that threshold at the time.
The presence of stronger unions was, as expected, highly correlated with districts selecting remote learning as the instructional plan.
But there are important caveats behind the findings. For one thing, the study, as with other recent research, is limited to school reopening during the fall鈥攍ong before a third national surge in COVID-19 caseloads. And the data found inconsistent results when rates exceeded the 36-to-44-persons-per 100,000 hospitalized rate.
鈥淭his possibility of increased hospitalizations is consistent with the idea that social interaction creates more negative health outcomes when there is more of the virus to be spread, perhaps despite careful school safety measures,鈥� the researchers note. 鈥淚n these cases, it could still be appropriate to reopen schools, because of the negative effects on students and parents from keeping schools closed, but there may be more of a trade-off.鈥�
Second, the authors note, the study looks at the district policy, not what would happen if all students returned to in-person schooling; nearly all districts gave students the option to remain in remote learning.
Still, the findings are generally consistent with That study, conducted by a team of researchers at the Education Policy Initiative Collaborative at Michigan University, estimated that in-person schooling in districts with lower test-positivity rates 鈥攂elow about 20 new cases per 100,000 people in Michigan and five per 100,000 people in Washington鈥攄id not significantly boost coronavirus rates.
But that study鈥檚 findings were also complex. It found that in-person schooling did contribute to transmission as community case rates began to exceed those figures. And, as with the Tulane study, many counties are now well beyond that threshold.
Douglas N. Harris, a professor and chair of the economics department at Tulane, and the director of REACH, noted that in December, only 58 percent of counties fell below the hospitalization rate of 36 to 44 people per 100,000.
And, he said, district leaders will have to take their local health figures into account, plus whatever researchers learn about the new variant of COVID-19.
鈥淚f it turns out that that variant spreads more easily and is just as harmful, then yes, you鈥檇 want to adjust the threshold down to account for that,鈥� he said. 鈥淲e鈥檙e in a fluid situation, and the research isn鈥檛 either fast enough or precise enough to account for all of that, and that鈥檚 where the judgment of decision makers will have to come in.鈥�
As the School Year Restarts, So Do Debates On In-Person Learning
The research isn鈥檛 likely to do much to settle the debate over whether school districts should have prioritized in-person learning last fall. But it could offer some clues about how to approach the rest of school year.
Whether or not schools should return to in-person schooling has supplanted testing, accountability, charter schools, and curriculum debates as the most divisive education policy issue of 2020, and appears likely to do so in 2021, too.
Especially in heavily populated urban districts, it鈥檚 unclear whether districts can find ways to bring more students back and ensure staff safety for the second half of the school year. Some districts and teachers鈥� unions continue to struggle to reach an agreement, delaying in-person schooling plans in cities like Washington and Chicago.
Districts have also faced a myriad of logistical challenges as they consider in-person learning, ranging from uncertainties over their liability to the challenge of social distancing on school buses鈥攁ll coupled with a near total absence of assistance from the U.S. Department of Education.
The impact of sustained remote learning is being debated, too. While learning loss is perhaps less dire than some researchers predicted, certain groups of students, such as those receiving special education services or those living in poverty, are likely to be harder hit academically by months away from their regular routines.
And while much ink has been spilled on the academic impacts of the pandemic, other advocates have pointed out the emotional impact of isolation on youths, as well as schools鈥� key roles in providing meals and reporting signs of physical abuse to child welfare agencies.
In some cities parent groups have advocated on both sides of the reopening debate, even as racial subtexts lurk just below the surface: Black families, who have long faced racism in both the K-12 and the health-care system, have generally been far more cautious about sending their children back to in-person learning than other families.
