Daily new U.S. COVID-19 cases continue to increase by about 50% per week.
Data source for this and other graphs of new case counts: Calculated from The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 7/21/2021, from https://github.com/nytimes/covid-19-data.” The NY Times U.S. tracking page may be found at https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html.
Crossing that fine line between arithmetic and fear-mongering.
It baffles me that I seem to be the only person in the U.S. who is well and truly freaked out by the current situation. And I think that’s because I can do the math, and I understand the data.
Based on what I’m reading in the newspapers, most people don’t quite grasp what compound growth rates can do. And nobody mentions that the case counts we are looking at today reflect infections that occurred roughly 16 days ago.
When you combine those two — rapid compound growth, and long data lags — that’s a high-risk situation. The lag between infection and reporting means that our governments are are reacting to the way the world was two weeks ago. If they only react when case counts reach crisis levels, the rapid growth means we’re bound to overshoot those crisis levels considerably.
It’s a trite-but-true analogy: This combination is like trying to drive at high speed while looking only in the rear-view mirror. Wrecks should be expected.
Let me now walk through that. If the current growth in new cases remains constant, here’s what we look forward to, in the next three weeks or so.
If new case growth continues at the current rate, just over one week from today, this fifth (Delta) U.S. COVID-19 wave will be the second-worst on record for the U.S.. And a little over three weeks from today, it’ll be the worst.
That’s not fear-mongering. That’s just arithmetic. And that’s the entire span of time available to us, to get this under control.
You want some fear-mongering? OK, the growth rate for roughly the next 16 days is already locked in. There’s not a thing we can do about it. Why? Because those infections have already occurred. Infections occurring (on average) today will fully enter the data about 16 days from now. It takes that long for the entire sequence of events to play out, and have those infections enter the data. That sequence is infection, symptoms, care-seeking, testing, reporting and finally seven-day-moving-average-smoothing.
And, as a corollary, the data we’re looking at today reflects infections that occurred, on average, 16 days ago.
And so, given that nobody seems to have done much of anything to address this current wave, the graph of projected new case rates probably actually looks like the picture below. If I had to guess, I’d guess that 40 new cases / 100K / day is already locked in.
OK, I was just kidding. That’s not really fear-mongering, that’s once again just arithmetic. Just an illustration of the magic of compound growth.
You really, really want some fear mongering? Keep in mind that I’m only talking about what I think is already locked in.
Per Post #1178, this latest variant still kills about 1.5% of the persons diagnosed with it, same as the prior variants. (Deaths lag new cases by about two weeks, so we’re not really seeing that yet in the daily numbers. Just give it a bit of time.)
Just for the fun of it, how many deaths per day does that 40 / 100K / day imply? Once again, that’s just arithmetic. It implies (330,000,000 * (40/100,000) * .015 =) ~ 2000 deaths per day.
That might be a slight over-statement, based on historical data. Then U.S. third (winter) wave peaked at 69 new cases/ 100K/day, and the deaths peaked at around 3600/day. That’s roughly the same ratio as 40 /200k/day and 2000 deaths calculated independently here. But, given that a high proportion of elderly are now vaccinated, if Delta were only as virulent as the then-prevalent (native, pre-Alpha) strain, we’d have to expect significantly fewer than the calculated 2000 deaths. On the other hand, if Delta is more virulent, that logic would not apply, and all bets are off.
To put that in perspective, in a normal year, about 2.8M U.S. residents die. That’s a rate of (2,800,000 / 365 = ) ~ 7500 deaths per day. There’s a bit of seasonality to that, but not much. It’s very close to a steady rate. So that will increase daily U.S. deaths by about 27%. And as we’ve seen countless times in this pandemic, the relevant supply chain is designed for dealing with normal demand and not much more. Given that this will hit unevenly, that additional 2000 deaths will definitely reach the “stack the bodies in refrigerated trailers” zone in some parts of the country.
How many hospitalizations does that imply, again relying on Post #1178? That’s (330,000,000 x (40/100,000) * ~0.085 = ) ~11,000 new hospitalizations per day.
Again, that might be a modest overstatement. That hospitalization number based on national data is significantly higher than what I calculate using more reliable data for Virginia. And, as with deaths, we’ve vaccinated a large fraction of the high-risk population. On the other hand, there are no guarantees that Delta is only as virulent as the native (pre-Alpha) strain of COVID. So I’m a little unsure that hospitalizations will reach quite that level.
Again, for perspective, in a normal year there are about 36M admissions to short-term general non-federal hospitals in the U.S. That’s a rate of (36,000,000 / 365 = ) ~100,000 admissions per day. But the average hospital admission lasts just 4.6 days, per the U.S. AHCPR HCUP, while the average COVID-19 admission lasts (as I recall) just about 12 days. So that 11,000 new daily COVID-19 admissions would increase the demand for hospital beds by ((11,000/100,000) * (12/4.6) = ) ~ 30%. Given that this will hit unevenly, some areas of the country will definitely see crisis-level shortages of beds.
On the bright side, maybe the virus will have started running out of bodies some time in the last 16 days. But a little bit of math suggests that’s unlikely. At (say) 20 / 100K / day average new infection rate for the period, and (say) three unreported infections for every reported infection, 16 days’ worth of infections would amount to (16* 20/100,000 * 4 = ) ~ 1.2% of the U.S. population. Given that half the population (40% of the adult population) is not fully vaccinated (per CDC today), I’d say it’s unlikely we’re running out of cannon fodder any time soon.
Just one more statistic, for now. What was the mean duration from onset to peak, for waves 2, 3, and 4?
- Second (summer) wave: 6 weeks.
- Third (winter) wave: 16 weeks.
- Fourth (Alpha variant) wave: 4 weeks.
- Fifth (Delta variant) wave: 3.5 weeks, so far.
We know why the fourth wave stopped. That was stopped by vaccination. That’s what the basic numbers suggested would happen (based on the infectiousness of that variant, Post #1051), and that’s what did happen. And the same arithmetic told us that’s not going to happen with this wave (Post #1160, Post #1173).
My point being that it this Delta wave merely lasts six weeks, as the U.S. second (summer) wave did, at this growth rate, it’ll end up being far worse than the winter wave of COVID-19. I’m not even going to show the new case rate six weeks out, based on the current rate of growth, because I don’t think it’s credible we’d get to that point without taking drastic action.
I should just end on that cheery thought. But I want to make two more observations.
First, check your toilet paper supply. (No, just kidding, that’s not even really funny, given the context. But FYI, yeah, I’ve checked mine.)
First, this is now a pandemic of the un-vaccinated. By and large, the behavior of Republican men and Republican governors will determine the course of this fifth (Delta) wave. Much of that population seems absolutely, totally committed to do nothing to stop the spread of COVID-19. This is a population whose creed requires no vaccine, no COVID-19 hygiene, and no restrictions on public gatherings, making this, in effect, a pandemic of the least cooperative and least public-spirited. The fate of this wave lies almost entirely in the hands of people who are sworn to let COVID-19 run its course.
Given that, what, exactly, would you expect to happen? My best guess is that this wave of the pandemic is going to run its course. As I said yesterday, my guess is that it will only stop only after the hospitals are full, because that’s the point at which Republican governors have “cover” for doing what their own public health experts have been telling them needs to be done.
But reasonable people might disagree about that.
Second, have you noticed how, suddenly, a lot of big-shot Republicans are finally promoting vaccination?
Have you guessed why? I have. There’s nothing hard about the math I’ve been doing above. Every state public health department is capable of doing the same. My guess is, those public health departments have informed their relevant state governments of what’s coming.
And suddenly, a lot of politicians who were doing their damnedest to screw up any effective response to COVID-19 now want to be remembered as having gotten on the right side of this.
But doing something effective — imposing mask mandates, say, or shutting down indoor public gatherings — would rile their uber-ignorant base. By contrast, preaching vaccination to a population that isn’t going to get vaccinated under any circumstances inconveniences no one. It does nothing other than address the tarnish on the reputations of those who have so suddenly and seemingly inexplicably gotten religion on this subject.
As I see it, the current set of about-faces among Republican politicians is not a case of finally seeing the light. Mostly, they’re plenty smart enough to understand the benefits of vaccination, and I’d bet that every one of them is vaccinated, along with their immediate families. So this conversion isn’t based on some sudden new understanding of the facts. Instead, it’s just the path of least resistance, given what they are now coming to realize is likely to happen next.
In fairness, Virginia has a Democratic governor who is both rational and a physician. Up to now, his COVID-19 response has been exemplary. And while Virginia’s seven-day moving average is only about 5.5 new cases / 100K /day, we’re seeing the same rapid rate of growth here as all the other states are seeing. For the last two days, the raw daily count has been over 8/100K/day, so that moving average is going nowhere but up for now.
Our governor is still sitting on his hands. And I have no idea why.