Back in summer, many school districts made a “deal” with parents and children. If you put masks on your kids, and allow us to severely restrict their ability to interact with other children as human beings, they can go to school. Parents were desperate, having watched the learning loss and depression of the spring, so they acquiesced.
This is part 1 of a 4-part series dedicated to un-masking children. Jump to other parts below.
Part 2: Understanding Relative Risk
Part 4: The Plan to Un-Mask Children
The science on masks was, and is not settled (in fact, prior to March 2020, it was settled in the opposite direction), and much of the health policy in this regard relies on specious research and conclusions. Children do not appear to be major drivers of transmission. Thus, regardless of whether or not masks “work,” we would expect limited transmission within school. Indeed, many have misused this as an argument for why masks work elsewhere.
Because many within communities, and particularly schools, believe that it is the masks and social distancing that are keeping them safe, they are afraid to give them up, no matter how great the long-term harm to children—we must provide a plan to allow children and those adults who want to, to return to normal. Ultimately relying on un-sound science has negative consequences. Universal masking is not an effective tool for control of a respiratory virus, but it does create a universal false sense of security.
In order to get out of this rut, we must instead look at the data. From there, we must develop a plan that acknowledges what we know and what we don’t, provides actions that will allow those who are concerned with the risks to remain a part of our community until they feel comfortable, and allows other to return to those aspects of life that they choose to, as they choose to.
To do these things, we must first get a handle on three key things:
Children’s role in spreading COVID, and triggering severe disease in at-risk populations.
Relative risk by age group, co-morbidities, and ethnicity.
The effectiveness of masks in controlling respiratory viruses.
From there, we can formulate action plans, based on this data, which will allow parents and staff to make sound, data-driven decisions that reflect their own personal risk tolerances, as well as an understanding of the level of exposure they have had heretofore. Let us turn first to children’s role in COVID-19 transmission.
Part 1: Children’s Role in Spreading COVID-19.
Children do not appear to be major drivers of transmission of COVID-19. This study is by far the most authoritative article on describing transmission networks. The lead author notes, they found not one incident of child-to-adult transmission. The study was a population-representative survey, in which they identified 600+ SARS-CoV-2 positive people who were included in the study. From this point, they used genome sequencing and contact tracing to identify how the disease spread through the community. This level of rigor is unique. Other studies that attempt to claim significant child-to-adult transmission of COVID have not performed this analysis. The genome-sequencing is what allowed them to say definitively that none of the cases was spread from a child to an adult.
Another study from Korea of 107 pediatric cases and 248 household contacts found not one instance of a child infecting an adult. The one example of a child infecting anyone was a 16-year-old child infecting a 14-year-old sibling—both parents were negative. This study is interesting as well, because the median age of the infected children is quite high. At 15, these would be children we would expect to more readily pass on the disease.
There are other studies that purport to show higher rates of child-to-adult transmission, but all rely either on faulty end-points (e.g. do not look at whether contacts were actually infected by a child) or flawed methodologies (broken down in excruciating detail here).
Results from 14 randomized controlled trials do not, and have not, supported universal masking as an effective means of disrupting transmission of respiratory viruses. In early April, a data-free scientific consensus for masking was arrived at--in direct contradiction to the large body of extant scientific evidence, and with no new evidence to support it. On April 3rd, Dr. Fauci recommended that all Americans should start wearing masks (after saying on March 8th, that they should not). He has since testified before congress saying that he did not recommend wearing masks earlier, because we needed to preserve them for healthcare workers. This is clearly untrue, as he recommended people start wearing masks on April 3rd, just as the spring surge was beginning, and medical mask scarcity was at its highest.
Once this data-contradictory consensus was arrived at, the scientific community began churning out studies to try and “prove” what had been disproved for more than a century. The studies reflect those strong biases, and many (like this one) have been retracted as a result—others ought to be, but have not. In fact, the benefit was so “decided” that few, if any of these studies, have an actual non-masked control group. The result is that, particularly when it comes to schooling, nearly all of the scientific literature focuses not on whether masks are actually useful in stemming the spread within schools, but on how to get children to accept masks.
Sweden offers a unique exception. It cannot help us with understanding transmission, because there was widespread community transmission. But it can help us to understand the health results of students and parents. Sweden did not close schools during their spring or fall wave. Nor did they require that children or teachers wear masks. The results of a country–wide analysis were published on 2/18/21.
There were two key findings that emerged from the study. The first was that the deaths in Sweden’s school children did not increase in the four months of the study period relative to the prior four-month period (or any other period). This is despite children being in school, un-masked, and generally living their lives. The other important finding was that nationally, fewer than 10 preschool teachers, and 20 schoolteachers who contracted COVID during the period received intensive care—none died. The study made a sex- and age-adjusted comparison, showing that the relative risk to teachers versus other professions was, in the case of pre-school teachers, 1.1, and for other teachers, 0.43—i.e. roughly the same for pre-school teachers, and less than half for other teachers relative to the general population.
This is in-line with other findings that have recently started to emerge showing that spending time with children not only doesn’t put adults at greater risk, but is actually protective. A recently-released study in the U.K. of over more than 12 million adults, shows that while people who lived with children were at no higher risk of contracting COVID-19, for those under 65, they were 25% less likely to die. For those over 65, there was no difference in the outcomes. The study authors believe that the mechanism for this protection is via cross-reactivity from exposure to prior coronaviruses introduced into the home by children (note, this is also one of the prevailing hypothesis on why Asian countries, where most coronaviruses originate, were not hit as hard by COVID-19). It is worth noting as well, that the U.K., despite multiple extended, national, police-enforced lockdowns has a higher COVID death rate than the U.S.
Figure 1
Source: The COVID Tracking Project
Several recent studies have shown that when in-school transmission does occur, teachers (who are masked) are central to those transmission networks. A recent CDC report investigated nine clusters across 6 school districts in Cobb County, Georgia. Of those 9 clusters, 8 involved a teacher (See figure 2 below). In the one cluster where a student was the sole index case, the student only infected other students. In half of the other 8 clusters, the teacher was indisputably the index case. In the other four clusters, researchers were unable to determine whether the student or the teacher was the index case. This once again gives further credence to the idea that children are not major drivers of spread. Of further note, all of the children were masked all day, except during lunch, which was taken in the classroom. Investigators reported that both reported and observed mask compliance were high. Though in five of the clusters, interviews indicated that mask use was at times sub-optimal. In the only cluster where a child was the sole index case, mask compliance was reported to be high and correct, even in interviews. Of note, in the next study we will examine, in Germany, while children over 10 were required to wear masks while out and about “on-campus,” no children of any age were required to wear masks in the class room.
Figure 2
Source: CDC
Another recently released study examined in-school transmission dynamics in the German state of Rhineland-Palatinate, a state with a population of 4.1 million, 1492 schools, and student population of 406,000 school age children, and 144,000 children under 6 in day care centers.
The study was undertaken from August 17-December 16th, during a period of exponential case growth. During that time, students were in school in-person, 5-days a week. Under 10, neither masks nor physical distancing was required. Over 10, masks were recommended on campus, but not in classrooms. Physical distancing of 1.5 meters was also recommended for children over 10.
During the study period, 74,733 COVID cases were recorded. Of those, 16% were in children, roughly equal to the percent of children in the population (18.3%). Of these 9700 cases, 1954 of these cases had information pertaining to school—84% were notified and asked to participate in the study.
Ultimately, the study enrolled 591 students and 157 teachers. Contact tracing determined that transmission from these teachers resulted in 169 new cases—roughly 1 new case for every teacher case. Of these teacher-sourced cases, roughly half were to students, and half to other teachers. Students, which accounted for 591 cases, resulted in additional 145 cases (roughly one new case for every four student cases). Of these student-sourced cases, 120 were to other students (83%), and 25 (17%) to teachers.
Table 1
This table above shows the transmission patterns. Basically, transmission from teachers is 4x higher than from students. When a teacher caught COVID, they were 14x more likely to transmit it to another teacher, than a student who contracted COVID would be. A teacher who contracted COVID was also 2.5x as likely to transmit to a child, than a child who contracted the disease.
During the study period, as infections increased, masking recommendations changed. On November 2nd, children over 10 were also recommended to wear masks in the classroom. It does not appear that this had any impact on case growth within the classroom—with November seeing the highest number of cases. Likely the curve simply reflects the seasonal arc of the disease.
Table 2
The American Teachers Federation claims that 530 teachers have died since the pandemic began. Edweek puts the number at 233. Both numbers include large numbers of teachers who died during the periods when schools were not in session, as well as in districts that have remained fully remote, providing yet another indication that community spread is the real issue.
Taking these numbers, we can back into an estimate of the lethality of COVID for teachers. As of 12/31/20, the CDC estimated that there had been roughly 83 million cases—roughly ¼ of the U.S. population. Case growth has not abated since then, so estimating that we are now at 1/3 of the population is a fair guess.
There are roughly 3.6 million teachers in the U.S.. If we assume that teachers have been infected at roughly the same rate as others in the community that would suggest that 1.2 million teachers had been infected. With 230 deaths, that is equal to an Infection Fatality Rate (IFR) of 0.02%. This IFR is identical to the IFR observed for flu in this age group in the 2017-2018 flu season. It is slightly lower than the national estimate for COVID deaths for 18-49 year olds of 0.03%. Many of the teachers who died fall into older age brackets, so we would actually expect this number to be higher, not lower. This is yet another indication that, as the U.K. study suggests, teachers are protected, rather than put at-risk by their profession.
Table 3: U.S. COVID-19 Deaths, Hospitalizations, Estimated Symptomatic Illness & Infection Fatality Rate by Age through 12/31/21
Looking at this another way, in an average flu year, roughly 10% of the 18-49 population has symptomatic flu. Of those, 0.02% die. If we assume that teachers are just as likely to be exposed to flu, this would mean that on average 360,000 would contract the flu, and 73 would die. This number, 73, is clearly lower than the 230 number observed to have died of COVID. However, there are two important caveats. First, we have actually observed two COVID seasons now, not one. For two flu seasons, we would expect 146 deaths. We have seen 230, according to EdWeek. Again, this includes many people who are older than 50, whose risk of death from flu is 3-20x higher than those under 50. Further, COVID is a novel disease. Thus, more people are expected to contract the disease than flu. As of now, as noted, roughly 1/3 (~33%) of the country has contracted the disease, more than 3x the 10% we would expect in this age group in a typical flu season, but again, this has effectively been over two seasons, not one. That cases are now dropping like a rock at this level (roughly 33%) was actually what was expected by many, more sober epidemiologists—who got less TV time than more prominent ones.
The totality of these studies shows the following:
Children are not a driver of transmission in schools (or anywhere).
Masking children in schools does not appear to impact transmission within schools—it is low regardless of whether children are masked are not.
Teachers and others who spend significant amounts of time with children appear to be at reduced risk of death from COVID, due to prior immunity conferred from their time with children.
By the numbers, teachers are not dying at higher rates than other professions—in fact, they appear to be dying at lower levels, levels which are exactly in-line with their relative risk for flu.
Teacher acquisition of infections appears to be independent of their being in school, given that many, if not a majority of teacher deaths resulted when schools were not in session.
It is a wonderful thing that children in schools do not appear to be a major driver of transmission or mortality in or outside of schools. However, even if they were, we would still be obligated to teach our children in-person. Children ARE major sources of transmission for the flu, and the flu is significantly more deadly to them, and roughly equally deadly to their teachers. While school children are not a major driver of COVID transmission to their grandparents (as demonstrated by the UK study), they ARE major drivers of flu transmission to their elders—even with vaccines (65% of school children and 65+ are vaccinated annually for the flu—the highest levels of any age groups). Yet for time immemorial, we have sent kids to school, and grandparents have gloried in their ability to spend time with their grandchildren. This should not change.
These facts demonstrate that many of the problems we have had coming to terms with COVID are related to our inability to put the risk into perspective, looking at relative, rather than absolute risk. The next post in this series, here, examines this.
Comentários