co authored by Nan Fulcher and Justin Parmenter
This week state health agencies presented information to the North Carolina State Board of Education regarding COVID-19 and schools.
Representatives of both the North Carolina Department of Health and Human Services (NCDHHS) and the North Carolina Pediatric Society made the case that the risk of children spreading COVID-19 in schools is minimal and must be balanced against student needs that can be most effectively met with schools open.
The presentation offered a potential preview of a school reopening announcement that Governor Roy Cooper is expected to make in the next few days.
However, the data that NCDHHS is using to drive this important conversation doesn’t necessarily support the conclusion that opening schools is safe for our children, and grim anecdotal evidence emerging from around the country suggests the actual risk of COVID-19 transmission in schools may be much higher than currently predicted.
During this week’s state board meeting, State Health Director Dr. Elizabeth Tilson advised board members that “Schools have not really seemed to play a major role in transmission” and talked through this series of bullet points downplaying the possibility that North Carolina’s schools could become a hotbed for COVID-19:Navigating-COVID-19-Schools-primer-3
The original source of the information on this slide is the Massachusetts Initial Fall School Reopening Guidance document, which was released last month as that state’s plan for the 2020-21 school year. The evidence for that document’s statement “schools do not appear to have played a major role in COVID-19 transmission” is based on the following studies:
1) In France, a February study of one of the very first accounts of COVID-19 in Europe, involved a nine-year-old child who went to three different schools following infection by an adult and none of the 115 contacts at those facilities contracted the virus.
2) In Ireland, 3 children and 3 adults who tested positive for COVID-19 at different schools in March didn’t infect any of the 1,155 total individuals with whom they had some sort of contact.
3) In Australia, evaluation of the 863 contacts encountered by 9 students and 9 teachers at their corresponding primary and secondary schools in April showed that only one other student became infected.
4) A review article from early May, analyzing COVID-19 case clusters. For example, a cluster would consist of an infected person as well as anyone else who got infected directly by the individual or secondarily via the individual’s network. The study examined the location of transmission and showed that only 4% (8 of 210) of the clusters involved schools.
This small collection of studies provides minimal evidence for the dynamics of COVID-19 spread in schools. Thus, the NCDHHS summary statement that “schools do not appear to have played a major role in COVID-19 transmission” may convey a more definitive conclusion than warranted, especially to those unfamiliar with the data.
A closer analysis of the four studies reveal flaws that further call this conclusion into question.
One pitfall has to do with the reason there are so few studies about viral transmission in schools in the first place.
When the virus started to spread in the Spring of 2020, schools were quickly closed. The only studies about the virus in schools come from public health teams that were able to document cases just as in-person classes were being cancelled. And once schools closed, the chance to study COVID-19 in that setting disappeared.
The fact that schools weren’t open during the majority of the study period affects the interpretation of this research. In the case studies from France, Ireland, and Australia, infected people were pulled out of school immediately, disrupting transmission. Of course this is proper procedure from a public health standpoint, but it might not reflect what would happen in every school going forward. For example, if an infected subject didn’t seek testing or quarantine as rapidly, the transmission rate could be much higher. There is also the potential for spread by infected, but asymptomatic, individuals.
Another consideration is that when these studies were being done in the Spring of 2020, SARS-CoV-2 was truly exotic, with single cases (typically acquired by travel) dropping into an unsuspecting community.
Today, after months of full-on community spread–not to mention the alarming rates of infection in many areas of the U.S.–it’s more likely that schools will be contending with multiple cases simultaneously, and on a continuing basis. These factors would almost certainly cause higher transmission rates compared to those from the earlier reports.
The other NCDHHS bullet point “If infected, children may be less likely to infect others with COVID-19” is also undermined by school closings.
In the three reports cited for this statement, pediatric cases from China, Switzerland, and the U.S. (Chicago) were being collected just as schools were shutting down. Thus, the conclusion that children weren’t spreading to other children was skewed by the fact that routine contact with their schoolmates and playmates was no longer occurring. Instead, children were isolated within their family group, which may or may not have included other children.
Even the results of the cluster analysis are likely to be skewed because it’s not a fair comparison.
Infection clusters can’t develop in schools when the doors are locked. In contrast, activities like in-person church attendance, which is still going strong in some places, can spawn new clusters. Thus, the number of religious gathering clusters would be over-represented in a survey like this, undercutting the proportion–and therefore underestimating the significance–of school clusters.
It’s also important to note that, in identifying the eight school clusters in their study, the authors actually prove that schools CAN be a focal point of viral spread.
One of the clusters cited is the dramatic case associated with the Salanter Akiba Riverdale school in New York, where over sixty people including staff, students, and family members were infected, although some transmission may have occurred via events outside school.
Another citation is from Singapore, where, despite the heroic measures of enhanced hygiene, staggered classes, temperature checks, and a week-long break away from school, there were still eight reported cases before the school was closed eleven days later.
This set of clusters is just a small collection of anecdotes, but they may be instructive for considering the spectrum of scenarios that could play out in our public schools.
Studies about school-based COVID-19 spread are paused for now, but anecdotal evidence from other places where children gather may reveal what’s on the horizon when schools reopen.
Recent reports of skyrocketing infections at daycare facilities (Texas and Charlotte, NC, for example) and the summer camp outbreaks in Missouri and Arkansas do not bode well for North Carolina schools filled–even partly–with potential carriers of COVID-19.
In summary, the available data for COVID-19 spread in schools does not present an open-and-shut case of guaranteed low risk for students and teachers. Instead, a more thorough reading of each case reveals there’s likely to be enormous variability in the way COVID-19 disease spreads within educational facilities, depending on any number of physical, social, and biological factors – factors we do not understand and therefore can’t predict with any degree of certainty.
As Dr. Anthony Fauci recently told members of Congress, “We don’t really know, exactly, what the efficiency of spread is” among children.
Public school stakeholders need to be aware of the pitfalls associated with these studies as they draw conclusions from the COVID-19 information supplied by NCDHHS. Policy decisions affecting the health and well-being of school children should not be based on flawed reasoning and studies that don’t accurately model the current patterns of infection in this rapidly evolving crisis.
Nan Fulcher earned her Ph.D. in Microbiology and Immunology from the University of North Carolina, specializing in infectious disease research. She’s involved in science and outdoor education programming for children and does freelance graphic design.