"Man, we've gotta open the gates!"
"We've been through this, Phil. The Mongols are still outside."
"But I've gotta harvest my turnips."
"MONGOLS"
"Ugh. But we've been in here for WEEKS."
"That's how sieges work, Phil."
"But the Mongols have barely killed anyone in days."
"That's because of the walls, Phil."
"Are you sure? Maybe the Mongols aren't that dangerous. I'm just saying, how bad could it be? They can't kill ALL of us"
"That is literally the thing they do."
"But my turnips..."
(from somewhere on the Internet)
This past Spring, I conducted 99 ‘remote learning’ surveys (parents, students, and teachers) for close to 50 independent schools. Forty-four of these surveys involved input from more than 13,000 parents. Across every surveyed school, support for adoption of remote learning as an interim measure was overwhelming (source). I’m more than a little intrigued to learn what has changed since early June when the last of these surveys was completed. It’s early to say but one thing is for sure. Fatigue is setting in and parents are much more mixed in how they view anything but full-time in-class learning. Fatigue has a way of changing positions. Facts are considered differently. Evidence is more easily dismissed. People believe what they want to believe, in the face of mountains of evidence to the contrary.
Politics aside, folks want their kids back in school. Of course, everyone wants their kids back in school. The whole world wants kids to be in school, myself included.
(Our youngest, at 21, is completing her fourth year of university entirely from home. We’d love to see her on campus, but fully support the at-home alternative. We’re blessed not to have school-aged children, and I feel for those who do. Alana has fully embraced the remote learning environment, though, and has assumed an active role on student council, advising younger students on how to navigate their way through a world of uncertainty. This probably won’t surprise you, but as a member of student council, she’s also successfully conducted a survey of 360 students on their experiences with remote learning. In sheer volume, her 100 slide PowerPoint presentation of results rivals my own work. Who would have guessed? Is there a succession plan for me in here somewhere?)
We’re in the middle of a meandering world-wide experiment, making imperfect decisions, informed by evolving and imperfect information.
I won’t try to list the many reasons people offer for going back to full-time, in-class learning. They’re all good reasons, and I will not dispute any of them.
What I will say, however, is that efforts to diminish the risk of the Coronavirus may prove fatal… literally. From where I sit, this has already taken place.
We don’t know enough to draw firm conclusions on the Coronavirus risks related to students being in classrooms full time without social distancing. History, alone, will judge. I won’t say that I know anything on the matter. Like everyone else, I have more questions than answers. What I will say is what I see. Here’s what I see.
Many are quick to declare that teenagers are highly unlikely to fall ill from Coronavirus. This is true. The mortality rate for those teenagers who do fall ill is very low. This is also true. Here’s the catch. It would be incorrect to conclude that a student who has not fallen ill has not contracted the virus. On its own, this represents an unseen risk to more vulnerable members of the community.
“Taken as a whole, the evidence regarding the prevalence of COVID-19 among AYA [adolescents and young adults] and their role in the forward transmission of SARS-CoV-2 in the United States is not conclusive, but data suggest current estimates may inadequately reflect the contribution of AYA to accelerating and sustaining COVID-19 outbreaks.” (source)
For me, this excerpt points to a big unknown that may represent significant risk to the community in general.
Because of their greater number of ‘close contacts’, teenagers are seen as possessing enhanced risk for forward transmission of the virus. One study suggests that this group may, indeed, be the “fastest spreader of the Coronavirus”. (source)
With an estimate of 40% asymptomatic transmission, we are left unprotected from an unseen source.
If, as suggested, there is no lesser chance for a teenager to contract the virus than for the general population, and no lesser chance for a teenager to spread the virus (perhaps even greater), everyone who comes in contact with an infected teenager (whether that teenager is symptomatic or not) is at risk for forward transmission. My point? Teenagers cannot be considered in a vacuum, apart from more vulnerable members of the community. Teenagers have parents, grandparents, teachers, and so on. If teenagers are just as likely as others to transmit the virus, and likely to do so without falling ill, and more likely to have broader social interactions, should they not be considered as key potential drivers of viral expansion?
There is also research suggesting that, for teenagers, there may be a shorter-than-average period between contracting the virus and sharing the virus with others. Concerning evidence continues to accumulate.
I’m a numbers guy. Numbers don’t lie. I lean heavily on science and scientists. As I saw online the other day, “working things out is better than making things up”. What follows are lots of numbers… I apologise… but these numbers are critical to understanding what’s going on, at least what I see is going on.
The truth is that COVID-19 numbers frighten me greatly. I apologise in advance if these numbers also frighten you. I’ve crunched and crunched and crunched numbers until my calculator finger is at risk of losing its print. I’ll try to keep this as simple as possible.
To get to the calculations, four definitions will help:
* case mortality rate - this is the percentage of people with confirmed diagnosis who are likely to die
* R-naught - the re-infection rate (written R0), or how many other people are likely to become infected directly from one person
* serial interval - the time in days between the onset of symptoms for one person and the onset of symptoms for the second person, infected by the first. I refer to this as the infection cycle.
* mitigation - steps taken to prevent the spread of a virus, including hygiene, social distancing, masks, and where available, vaccinations
Everything is relative, of course, so let’s compare the Coronavirus to the common flu. After all, that’s the comparison put forward by so many people.
The common flu numbers for the terms I’ve described here are well established. We have a long history of experience to work from.
It will be at least a couple of years before any certainty presents itself for these figures as relate to the Coronavirus, but many independent studies have already generated ball-park estimates. Are they correct? We cannot know as yet. Time alone will judge. I like the saying, “All models are wrong, but some are useful.”
For the sake of making a conservative argument, I will employ very optimistic estimates (from within research ranges) of these factors in my calculation. If I were to use more recent estimates for a couple of these variables, the picture would be even more frightening.
The case mortality rate of the common flu is generally accepted at 0.1%. That is, one-tenth of 1% of people diagnosed with the common flu are likely to die from related complications. For the Coronavirus, I’ll use a common estimate among the experts, at 1%. That’s still 10x as deadly as the common flu. Quite frankly, for me, even at 10x, this is the least scary aspect of the calculation I’m about to perform. Even if the case mortality rates were identical, the differences, as I will show, are still significant.
Here's the key to my greatest concern. The R-naught for the common flu is well known at 1.3. That is, on average, each person with the common flu will infect 1.3 other people, in an environment without mitigation. Estimates of the re-infection rate for the Coronavirus vary widely. Until recently, the range from expert models has it between 2.0 and 3.0., again, without mitigation. More recent studies suggest that this may actually be in a range as high as 5.0. For the sake of conservativism, I will stay low, using 2.2 as the R-naught for Coronavirus. It is this rate of re-infection that commands so much of the attention of health authorities, as it should. The re-infection rate must be driven below 1.0 by mitigation efforts before we can hope to enjoy success in this battle. Below 1.0, the virus will die out on its own.
With much still unknown about the Coronavirus, the infection cycle is thought to be as short as 5 days (for teenagers, by some estimates, as short as 2 days). In other words, from the first person’s onset of symptoms to that of the next, as few as 5 days pass. I’m going to go way out on a limb, just to be conservative, and peg this at 14. I’ll use the same interval for both viruses, though both are assuredly shorter, so that we can isolate the critical role of the re-infection rate, the contagiousness of the Coronavirus, all else being held equal.
So, where do we go with all these numbers? Let’s start by framing the comparison over a typical flu season, say about 5 months, or 20 weeks. Well, 20 weeks is 10 cycles of 2 weeks each, right?
One infected person with the flu, as discussed, will infect 1.3 people in the first cycle (2 weeks). Carry the math through 10 cycles in the season and we’ll find 13.8 people infected after 20 weeks in an unmitigated environment. This is simply: 1.3 x 1.3 x 1.3 x 1.3 x 1.3 x 1.3 x 1.3 x 1.3 x 1.3 x 1.3 = 13.8.
Now, let’s do the same with my conservative re-infection rate for Coronavirus: 2.2 x 2.2 x 2.2 x 2.2 x 2.2 x 2.2 x 2.2 x 2.2 x 2.2 x 2.2 = 2,656.
This means that, after just 20 weeks (assuming the 14 day re-infection interval, rather than the estimated 5 day span), for both the common flu and the Coronavirus starting from just one infected person, the number of people infected with the Coronavirus will be 192 times that for those with the common flu. This is what they call exponential growth. While 2.2 doesn’t seem like that much more than 1.3, small differences at the front end expand dramatically over time through multiple cycles of infection.
Now, let’s bring into the calculation the case mortality rate difference, at 10x. Now we’re looking at potentially 1,920 times as many deaths from the Coronavirus as for the common flu. Even if the difference is only 6x (which I suspect will be closer to the mark once the dust settles), we’re still seeing more than 1,000 times the number of deaths. Even if, as some claim, the mortality rate for the Coronavirus matches that for the common flu (which it absolutely does not), at 0.1%, you’ve still got 192x as many deaths, by virtue of the exponential growth in the number of infections. 192 times the average annual death count from the common flu... you figure it out. It’s too depressing to write it down here. In the USA, the average annual deaths from the common flu number about 40,000 (with variability from year-to-year... and that’s a virus for which a vaccine provides meaningful mitigation). In the absence of mitigation, my framework of thinking suggests a total number of deaths from the Coronavirus in the multiple millions.
This is why I say that the re-infection rate (the R-naught) is the centre of the matter, and this is why the Coronavirus scares the wits out of me. As I study the numbers, I’m guessing that our public health authorities are afraid to share the full magnitude of risk associated with the Coronavirus in an unmitigated environment.
Seven months in, the number of dead Americans from the Coronavirus matches the per capita number of dead Americans at the seven month point of the flu pandemic of 1918. Think on that for a moment. While current estimates have the 1918 virus more deadly than the Coronavirus on a case basis, the current virus is much more infectious (R0=1.8 versus 2.2; to be clear on this point, 10 cycles of re-infection at 1.8 yields 357 infected versus 10 cycles at 2.2 resulting in 2,656 infected). The risks are spread by this exponential growth to a far greater proportion of the population. The 20-week season of re-infection, as I have described, would see more than 7x the number of infections as for the 1918 flu, in an unmitigated environment. That difference more than compensates for the spread in case mortality rates. I will be surprised if, in the end, we don’t see more Americans die from COVID-19 than died from the pandemic a century ago. That number? 675,000! What’s more, in the absence of mitigation efforts, if the per capita equivalent for the Coronavirus holds, with triple the population from 1918, deaths could be over 2,000,000.
The operative phrase here is: in the absence of mitigation efforts. There are great masses of data supporting the recommended mitigation strategies. With varied levels of adherence around the world to basic guidelines, valid success rate comparisons are possible. In short, the basic mitigation strategies are indisputable for anyone who can read and think. As Adlai Stevenson declared, however, “we need a majority”.
21% of the American population is over the age of 60. That’s 68 million people. Are they expendable? Few of these people are in nursing homes, so how do we protect them?
40% of the American population is obese. That’s 130 million people. Are they expendable? How do we protect them?
Those in the high-risk cohorts include not just people over the age of 60 and the obese, but also diabetics, asthmatics, cancer survivors, and the list goes on. I count these groups as representing greater than 50% of the population in the USA. How do we isolate half of the population from the other half of the population? In the absence of the coming vaccines, all we have for now are: physical distancing (with efforts to retain critical social interaction), appropriate hygiene, and the wearing of masks.
In the end, leadership and consistent messaging will be key to navigating our way through this crisis. Leadership, in troubled times, is a very lonely place. Successful leadership during this period of uncertainty will hinge on helping us all to see ourselves as interdependent members of a community much larger than ourselves. Teenagers, as members of this community, cannot be separated and considered in a vacuum. Failure to understand this interdependence and failure to account for it in mitigation strategies involving our teenagers, can yield only failure.
… but my turnips!
With respect,
Kevin Graham