Don’t rely on the ABC Science Collaborative study to inform school re-opening decisions

As schools begin spring semester classes, local leaders in North Carolina face the weighty decision of whether to offer in-person instruction. Local leaders have been handed no favors from federal and state officials who have abrogated their responsibility to establish objective criteria for in-person instruction that would apply to all schools, public and non-public alike.

The advocacy of a vocal minority of disproportionately white and wealthy parents, coupled with competition from unregulated private schools and lightly regulated charter schools, are further pressuring school boards to de-prioritize student and employee safety.

As school boards have debated the issue, many advocates for a return to in-person learning – including the State Superintendent – have pointed to a new study from the ABC Science Collaborative. The study, led by physicians at UNC and Duke University, examined data from 11 North Carolina school districts to conclude that COVID secondary transmission of COVID in schools is “extremely limited.” As a result, the authors argue “schools can stay open safely in communities with widespread community transmission.”

On its face, the study is good news for those supporting an immediate return to in-person schooling. However, a close reading of the study shows that it offers little useful guidance for  determining whether a return to in-person schooling is currently safe. The study has several shortcomings: It almost certainly underestimates COVID transmission from earlier this school year, conflicts with numerous other studies, and was conducted under conditions that aren’t comparable to what schools are facing today.

The study examined 11 school districts over the first nine weeks of school from the end of August through October. Researchers relied on the contact tracing of local health department staff to determine whether they could track confirmed cases to within-school transmission. Among students and staff, the researchers identified 805 confirmed cases over the nine-week period. Of these, they estimated that just 32 were acquired within schools; they were unable to identify any cases of child-to-adult transmission.

Like many academic papers, however, the findings are not ironclad. In fact, the paper itself devotes an entire section to the study’s limitations. Unfortunately, these limitations have been summarily ignored by reopening advocates and most press accounts.

As the report notes, participating districts volunteered; since the sampling was not random, it could bias the results. Districts that volunteered for the study could have different health protocols in place than non-participating districts. COVID mitigation strategies in the study districts are possibly – if not likely – more rigorous than in the districts that are not working with the Collaborative.

Additionally, the report relied on the contact tracing efforts of local health departments rather than universal testing of students and staff. Anecdotally, there’s great reason to be skeptical of the diligence of local contact tracing efforts and people’s cooperation with them. The report would benefit from describing what contact tracing entails in each district, as well as the criteria for classifying an infection as one resulting from in-school transmission.

What the report tells us about these efforts is troubling: There was little testing of contacts, as testing was not required to return to school. Undoubtedly, the researchers’ decision to rely on contact tracing creates a high bar for identifying community spread and fails to identify a number of cases.

Finally, the report notes that even though these schools are considered “in-person” they were still operating under a hybrid plan under which roughly one-third of students continued to learn remotely. Even if you take their study at face value, it only argues for a return to “Plan B” rather than a full return of all students.

There are many additional concerns unmentioned by the report.

• Community conditions today look much different than when the study was conducted. Per the report, “across NC, the rate of SARS-CoV-2 infections was 1–2 cases per 1,000 residents per week for most of the 9 weeks examined.” As of Jan. 15, the county with the least number of cases over the past seven days, Watauga County, has 2.76 cases per 1,000 residents. In Hyde county, there have been a stunning 20.86 cases per 1,000 residents over the past week. Most counties appear to be in the 5-8 cases per 1,000 residents range.

• COVID looks different today than it did when the study was conducted. We now have new COVID variants in the United States that are reportedly 30% to 70% more contagious.
• The study ignores asymptomatic spread. By relying on contact tracing of confirmed cases, and by failing to require testing of contacts, the study ignores cases of asymptomatic spread. Over half of cases are thought to be spread by asymptomatic carriers. Children are more likely to be asymptomatic, yet still have the same capacity to spread the virus as symptomatic carriers.
• District personnel and enrollment data fails to align with state data sources. The figures for district enrollment and staffing differs from the data these districts report to the state. These discrepancies offer reason to question how data was handled overall.

Additionally, the ABC study is contradicted by many other studies and reports that find that schools contribute to community spread, particularly in situations like we are currently experiencing where community spread is high.

A Tulane University study found that schools likely contribute to an increase in COVID hospitalizations when area hospitalizations reach between 36 and 44 per 100,000 people per week. Right now, 40 of the 80 North Carolina counties for which we have data are beyond the 44 hospitalizations threshold, with another five exceeding the 36 hospitalizations threshold.

A second looked at data from Washington and Michigan and found that in-person schooling exacerbated community spread when there were between five and 20 new daily cases per 100,000 people. North Carolina finds itself at 79 new daily cases per 100,000 people.

These studies aren’t the only ones that find schools contribute to community spread, contradicting the findings of the ABC Science Collaborative. Here are 11 more. All of these reports pre-date our current high levels of community spread and the emergence of new, more contagious COVID variants— and therefore underestimate schools’ potential to contribute to community spread.

At best, one could describe the research on the question — “Do schools contribute to community spread” — as “mixed.” Some say yes, some say no (though having read the reports, I find the studies that say “yes” to be more persuasive). Given the risks, we should default towards exercising caution. Getting this question wrong risks not only death, but a risk of persistent, long-term health problems. Remote learning should be the default model until we have a higher degree of certainty on the safety of in-person learning, or, better yet, a vaccinated workforce. Epidemiologists agree with this approach: 75% surveyed said they would keep their own children out of school, and that was when conditions were much more favorable than they are now.

Not only are the risks of returning substantial, but the benefits are also questionable. In person-schooling during a pandemic doesn’t look like in-person schooling in normal times. Students are distanced, unable to work closely and collaborate in small groups. They can’t socialize or talk during lunchtime. In many cases, “in-person” learning is simply remote-learning taking place inside a school building instead of a home, with students sitting 6 feet apart, silently tapping their laptops.

Meanwhile, infections – whether acquired in school or not – lead to frequent disruptions as students and staff are required to quarantine, bouncing back and forth between remote and in-person. Even the 11 districts from the ABC study “reported quarantining more than 3,000 children and staff over nine weeks.” This can be more disruptive than simply sticking with a remote model and focusing on making it as effective as possible.

Others argue that returning to in-person learning is an equity issue since Black and Brown students are more likely to face barriers to remote learning such as lack of computer devices, internet connectivity, parental assistance, and the time and space to dedicate to remote learning. However, surveys consistently show that it is Black and Brown families who will disproportionately withhold their children from in-school learning environments. If in-person learning during a pandemic is as beneficial as its proponents claim, then a return to in-person learning is likely to exacerbate learning differences if Black and Brown students continue to learn remotely.

Anyone genuinely worried about educational equity would be better served focusing their advocacy efforts on delivering students the resources necessary for a successful school career. Parents and other stakeholders can learn more about these efforts via the Every Child NC website.

Ultimately, the preponderance of evidence indicates that districts should opt to keep their schools remote, lest they exacerbate community spread and sicken their community. The ABC Science Collaborative has provided schools with useful information on mitigation strategies, but their study on within-school transmission has glaring weaknesses and does not apply to our current situation of rampant community spread and new COVID variants.

If policymakers are truly interested in a safe return to in-person schooling, they must take steps to drastically improve community conditions (i.e., mask mandates, business closures, paying people to stay home), prioritize vaccination of school employees, upgrade school HVAC systems, and provide schools with the resources necessary to fully-support our students’ academic and emotional needs upon their return.

Kris Nordstrom is a senior policy analyst with the North Carolina Justice Center’s Education and Law Project.