A month ago, North Carolina public schools were finalizing their reopening plans. It felt as if something might finally return to normal.
Then a wave of new scientific studies began to appear, some of which suggested children were not as safe from COVID-19, the disease caused by the new coronavirus, as many initially thought.
After months of public health policy updates, controversy over masks and drug trials, and enough misinformation circulating online to cause an infodemic, these studies seemed to signal yet another drastic shift in our understanding of the pandemic.
Despite some ominous headlines, “really I think these studies have found what we had assumed all along,” said Dr. Ibukun Akinboyo, a pediatric infectious disease physician and infection control specialist at Duke University School of Medicine.
“My worry,” she said, “is around some of the interpretations of these studies. In some cases, we merge interpretations together and we start applying them to studies that are addressing different things. “
Two studies, one in the Journal of the American Medical Association and another in the Journal of Pediatrics, reported that young children were found to have high concentrations of viral RNA in their nasopharynx, despite presenting only mild symptoms.
Viral RNA is not directly representative of actively infectious viruses but is often used as a proxy, meaning that high levels suggest a potential for more efficient transmission of disease.
Another article, this one a contact tracing study from South Korea, found that children ages 10-19 had an 18.6% chance to also infect someone else in their household. Children in the 0-9 age group had only a 5.3% probability.
These two findings, high viral RNA load and documentation of child-to-adult spread, suggested to some that children may silently play a key role in driving community spread of COVID-19.
Akinboyo said the South Korean study and several other contact tracing studies confirm that children can transmit the virus.
“However, I would contrast the South Korean study with the CDC report on day cares in Rhode Island,” she said, referring to the national Centers for Disease Control and Prevention.
The Rhode Island study sought to determine how the virus spread when children were in contact with each other and adult caretakers. In this study, in most cases when a child was found to have COVID-19, there was no secondary transmission to other children or adults.
“I think this finding is very important,” said Akinboyo. “When everyone is complying with recommended practices that we know mitigate disease spread, we can have children who present with high viral loads and still limit transmission.”
The scientific process
These studies highlight the difficulty of building a scientific consensus in the middle of a pandemic.
There is no laboratory setting and thus no ability to conduct experiments while controlling for the many variables that affect transmission.
“You are pushed to get information out to the public as quickly as you can. However, that messaging is being done at the same time you are gathering information, which can lead to problems,” said Dr. Christopher Woods, a professor of medicine at Duke who has studied emerging global infections like Zika, HIV and Ebola for over two decades.
“The scientific process itself is fraught with steps forward and steps back,” he said. “It is imperfect.”
Science is a constant conversation full of dissenting voices and sometimes contradictory evidence. Researchers are used to existing in that world of uncertainty, but this year the scientific process has found itself under the bright spotlight of public view, and research is being conducted at an unprecedented speed.
“This is all happening so fast right now,” Woods said.
“Messages start to get blurred because of how quickly they come out. That is dangerous.”
Public health policy and communication
Dr. Leah Devlin, who served for 10 years as the state health director for North Carolina and now teaches at UNC Chapel Hill’s Gillings School of Global Public Health, also stressed the importance of clear and accurate messaging.
“In a time like this, of high risk and high concern, it is so important to have very frequent, public and transparent communication that is as accurate as you can be with the data you have,” she said.
Devlin explained that public health is a science unto itself. Health officials serve as the bridge between research science and public policy recommendations to government officials.
“Every policy decision needs to be grounded in the data,” she said. “If you have no data behind a decision, you have no footing.”
Problems arise when so much data exists that certain numbers can be selected that may misrepresent the overall picture.
“We have a lot of data out there, and our understandings are constantly changing,” Woods said. “You can start slicing those numbers.”
A recent example of inaccurate data reporting came last week when U.S. Food and Drug Administration Commissioner Dr. Stephen Hahn and other White House officials overstated the findings of a Mayo Clinic study on convalescent plasma.
At the press conference, officials said a plasma treatment reduced mortality by 35%. Hahn then inaccurately said that, with the administration of blood plasma, out of “100 people who are sick with COVID-19, 35 would have been saved.”
Many health officials were quick to point out that the 35% efficacy was overstated and based on a subset of a subset of people included in the study.
“This is another example of misrepresentation of data for policy purposes,” said Woods.
Woods further stated that he does not think the mistake was malicious but “likely opportunistic.”
“I don’t think the misrepresentation was deliberate,” said Woods, “but (it) demonstrated a weak understanding of the literature (at least by his staff).”
Hahn later acknowledged that the critiques of his statements were warranted and apologized for the inaccuracy, adding that the statements were not politically motivated but were based on convalescent plasma’s “potential to benefit many sick patients.”
“We need to focus on reporting accurate data,” Woods said.
“But we also have a duty to do so in a way that allays fear. Fear is the greatest problem we face in a pandemic … and it needs to be measured and appropriate.”
‘Children’ from 1 to 22
Returning to the studies on children and COVID-19, similar issues with data analysis and reporting could represent one reason behind conflicting findings.
For example, there is no universally agreed definition for “children” in research. Some of the studies sporting headlines regarding “children” as spreaders of COVID-19 include data from subjects up to 22 years of age.
This has important consequences because much of the debate around schooling and child care facilities has centered on elementary school-age children and younger.
Grouping 4-year-olds with 20-year-olds is not necessarily useful when deciding how to open schooling at different levels.
“It’s not as simple as children either are or are not infected,” Woods said. From masks vs. no masks, droplet vs. aerosol transmission and now in children, “these things are much more nuanced. It’s all about numbers, risk and likelihood of things, as opposed to binary yes or no.”
“Most of all, these studies highlight that we need more studies dedicated to children,” Akinboyo said.
Several studies show the rate of COVID-19 in children is strongly correlated with the overall prevalence in a community, she said.
“As adults, we need to do more,” she said. “Wear a mask. Model behaviors that we know mitigate spread. … If we get community prevalence down, reopening schools becomes a much easier discussion.
Click HERE for broadcast script.