The Michigan outbreak is now being characterized in national news as being driven by school children, and in particular, by school sports. That’s what showed up when I opened up the Google News website today.
Source: Google news, accessed 9 AM 3/26/2021.
That’s not implausible. If you look at college COVID-19 re-opening guidelines, student athletes have always been considered a high-risk class of individuals. They are a “high contact” population, in the jargon. (That’s not as in “contact sport”, but as in, they are going to have a lot of close contact with a lot of people. Probably maskless and breathing hard.) They are to be given particular attention in any college re-opening plan, including frequent testing.
Michigan was already reported as requiring weekly testing of student athletes. It doesn’t look as if they would even consider suspending high school athletics. So I guess they’ll just have to deal with it. Their rate of reported new case growth is high enough that it’s driving up the average for the entire Midwest.
Source: Calculated from: The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 3/26/2021, https://github.com/nytimes/covid-19-data. Their U.S. tracking page may be found at https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html.
But there are couple of odd things about this Michigan outbreak. Both in general, and in relation to school sports, which seems to have been chosen as the culprit.
Infectious COVID variants, or not?
First, it’s not clear that Michigan does or does not have a high proportion of cases with the highly-infectious COVID variants such as the U.K. B.1.1.7 variant. CDC provides contradictory evidence.
Per the CDC’s revised pages on coronavirus variants, Michigan doesn’t even make the list of states that have a high proportion of these new, more-infectious variants. The red circle indicates a lack of Michigan on the table:
And yet, an older CDC version of what ought to be more-or-less the same underlying data showed Michigan as a real hotspot.
If the older data were correct, then Michigan would fit right into the current narrative, which is that we’re now in a race between the more infectious COVID variants, and vaccination. A combination of a high rate of B.1.1.7 in the state, and low or no vaccinations in the young adult/school age population would dovetail the the stylized facts of Michigan as a) having an outbreak, b) concentrated among the young.
But the CDC evidence is contradictory. The more recent table shows no such issue in Michigan. (And in any case, the timing is wrong, as Michigan’s case increases started before they could plausibly have B.1.17 as a large share of cases.
So, this this in any sense fit the narrative or not? Based on the most recent CDC data on variants, and the timing of the start of the outbreak, not. (And yet, the older CDC data contradicts that.)
Huge testing artifacts have not been accounted for?
Second, as far as I can tell, having a school-sports-related outbreak of this size is rare. Vanishingly rare. We all know that mainstream media love to report about things that threaten children. But despite that bias, after doing several variations on internet searches, I found exactly one other mention of a sports-related COVID-19 spring outbreak, in Peoria, Il.
And while people are unquestioningly accepting that this is a school kid’s pandemic in Michigan now, focused on student athletes, nobody seems to be asking why. Plenty of states have in-person schooling in session now. Plenty have their high schools open. Many have baseline rates of COVID-19 infection that were about the same as existed in Michigan. Nobody is having this huge a problem with student athletes. Why, uniquely, are we seeing these cases taking off among school children and particular sports participants, in Michigan, but not elsewhere?
And so I have to ask the obvious question: What fraction of the apparent school-sports outbreak in Michigan is an artifact of intensive testing? My understanding is that a large proportion of COVID-19 cases in school children are asymptomatic, similar to what was reported in this study. You can also see news reports of vastly higher proportions.
Not to downplay the seriousness of it, but if you set about with mandatory weekly testing for some subset of the population, and that population has a lot of cases that are either asymptomatic, or so mildly symptomatic that they would not other wise be tested, then you’re going to see a big jump in diagnosed cases for that sub-population.
I have run into this problem with my ongoing analysis of William and Mary students. Those students are tested far more frequently than the Virginia average, and so, there is no truly good way to compare their apparent rate of COVID-19 infection with the Virginia average. But I took a stab at getting a rough estimate of the impact (Post #1032).
Bottom bottom line: With all the assumptions made here, you’d expect mandatory once-per-month testing to yield three times as many positives as you would get from voluntary, symptom-driven testing in the community. That’s because you’ll catch the one severely symptomatic case under either approach. And the mandatory once-a-month testing will also catch half of the four missing mildly-symptomatic or asymptomatic cases not found under voluntary testing.
That’s with a college-age population, and monthly testing. With school-age, and weekly testing, I would have to assume that the multiplier ought to be much higher.
In other words, if my William and Mary analysis is anywhere near right, Michigan should expect to see that their student-athlete population has many multiples of the diagnosed COVID-19 case rate as the rest of their student population, purely as an artifact of their mandatory weekly testing.
Given how large that effect is, I wouldn’t depend on the diagnosed cases to make an inference at all about whether or not there’s a problem with school sports in Michigan.
At the minimum, I wonder how state public health officials in Michigan are accounting for this in their analysis. And so far, I haven’t stumbled across anyone saying anything about it. So it doesn’t look like they’ve done any correction to the diagnosed case rate to account for this huge artifact of the enhanced testing of student athletes.
I tried to find some testing-independent indicators, such as COVID-19 hospitalizations in Michigan for the school-aged population. I couldn’t find the data.
There’s no doubt that Michigan is having a real outbreak among the non-elderly-adult population. You can read, for example, that hospitalizations are way up. And they may be having some sort of school-related outbreak. But if so, they appear to be more-or-less the only place in the U.S.A. where that’s occurring, despite resumption of in-classroom learning (and school sports) in a large number of states. I don’t see any public health discussion of the need to normalize the student athlete data for the mandatory and vastly higher rate of testing. And so I have to wonder how much of this school-sports-in-Michigan story is true. And how much is an artifact of the mandated resting regimen.
There’s actually a policy point in all of this. If the question of school sports comes up, and somebody points to Michigan as the reason for not resuming sports, just take that with a grain of salt. Michigan is definitely having some kind of problem. Whether school sports is the main driver cannot be determined from the raw rate of diagnosed cases in Michigan. That is too strongly perturbed by the mandatory weekly testing. And I don’t see anybody digging any deeper into the issue to determine the true story.