Expected Spread of Covid (Coronavirus) in Minnesota
People keep asking me how bad this thing could get.
I don't think it's helpful to look at the national level, because there is no single national outbreak. There are a series of local outbreaks, each of which is currently in a different stage. Washington State, right now, is under serious strain right now, desperately trying to mitigate what has already become an out-of-control epidemic. Minnesota has several cases, but (supposedly) it is not yet spreading person-to-person here, so Minnesota is still trying to work on containment, delaying the start of a real outbreak. Alabama has no cases at all, so it's simply in monitoring-and-preparation mode, hoping to recognize covid (and contain it, for a while) when it does finally show its face in Dixie. Unfortunately, Alabama cannot lend its hospital beds and nursing staff to Washington State; those resources are fixed in place.
So when I want to know how risky it is for me to go to church on Sunday, it's not that helpful for me to look at the number of cases nationwide. Most of them are thousands of miles away from me. I want to know how many cases there are around here -- and, more importantly, how many there will be a few days from now.
Something interesting: so far in the epidemic, diagnosed case counts have grown by approximately 33% daily, unless there is dramatic government intervention. Even more interesting, this has been fairly consistent at both the local and national levels.
For example: in the morning on March 2nd, the U.S. had 102 diagnosed coronavirus cases. Suppose you wanted to predict how many cases the U.S. would have in the morning today, March 12th. Assuming a 33% average daily growth rate, you would have pulled out your calculator and typed:
102 * ((1.33) ^ 10) = 1,766 cases expected by March 12th.
How many diagnosed cases did the U.S. actually have this morning? 1,697, off by 4%. Not bad, for an estimate made 10 days ago about a virus whose case count doubled four times.
Similarly, in the morning on March 4th, Seattle had 39 diagnosed cases. Assuming a daily 33% average growth rate, how many would Seattle have eight days later, on March 12th?
39 * ((1.33) ^ 8) = 381
How many did they actually have? 366, again off by 4%.
I played around with this looking at recent historical data for Italy, and Colorado as well, and it did pretty well. 33% growth per day has been a pretty decent average. Sometimes the estimate went a bit high, sometimes a bit low, but it was always fairly close to the truth.
I'm not really sure why this works so well. These are only diagnosed cases, and our testing regime is very bad right now, so we are certainly lagging the actual progress of the disease. At times, diagnosed cases may only be a fraction of real cases. (We saw that in Wuhan, although they eventually caught up.) So I wouldn't expect to see the diagnosed case count grow so smoothly. But, hey, it does, and I'll take it.
What I want to know is how bad it's going to get in Minnesota.
This is only an estimate, and not a very smart one. It doesn't account for any specific factors on the ground, community spread vs. imported infections, etc. And it only predicts diagnosed cases; there are likely many more undiagnosed cases, and we won't really know how many until we start having people die of it. Yet I'm going to stick with this formula, since it has proven fairly good at predicting the future, and I don't have anything more sophisticated at hand.
As of dawn on Friday, March 13th (oof, Friday the 13th, I just noticed that), we will have 9 diagnosed cases in Minnesota. (That's not an estimate, that's a news report.)
By dawn the following day, March 14th, assuming a 33% growth rate on average, we can estimate that Minnesota will have approximately 12 diagnosed cases.
By Sunday, March 15th, we can estimate 16 diagnosed cases.
By Wednesday, March 18th: 37 diagnosed cases.
By Sunday, March 22nd: 117 cases.
By Wednesday, March 25th: 276 cases.
By Sunday, March 29th: 863 cases.
By Wednesday, April 1st: 2,030 diagnosed cases.
By Sunday, April 5th (Happy First Contact Day):, 6,351 cases.
By Wednesday, April 8th: 14,941 cases.
By Sunday, April 12th: 46,751 cases... approximately 1,589 of whom will die.
By Monday, April 20th, we arrive at 450,000 diagnosed cases.
The day after that, the total number of cases exceeds 10% of the state's population. At that point, the virus will naturally begin to slow down, because enough of the population will have developed immunity (or died) to start to grant herd immunity. Covid won't actually stop, of course, but it will no longer grow on a 33% daily curve. A statistically significant chunk of Minnesota's over-60 population will already be dead or dying, but the deaths won't stop for some time after.
That's where we are headed right now, on our current trajectory. 9 (known) cases today turns into tens of thousands in a matter of weeks. That's exactly what happened in Italy.
There appear to be three ways to prevent this exponential growth in case counts from happening:
Stop testing. Can't have diagnosed cases if you don't try to diagnose! This won't reduce the case load or the fatality rate, and actually will make things much worse and kill lots of people needlessly... but it will make your numbers look good, so there's that.
Wait until the virus burns itself out (basically, until so many people have been infected that it can't infect enough new patients to remain an epidemic) and hope that it burns out much faster than other flu-like diseases. Maybe you'll get lucky, and Minnesota's wet, warming climate means that the infection rate will start to fall long before late April! There seems to be no reason to believe this will happen, but it COULD!
Impose and support major restrictions on the movement and gathering of human beings in Minnesota. Ban parades, ban sports, close schools (perhaps leaving them partially open, on a skeleton-crew basis, solely for the children of health care workers who have nowhere else to put them), cancel concerts, require businesses to support telecommuting and/or provide paid time off. Shut. Down. Everything.
Iran has tried option (1), and it's working out pretty well for them. Although a number of senior officials have died and they're currently digging mass graves, their official numbers aren't growing quite as fast as you'd expect them to grow if they were doing honest testing and reporting. And that's the goal of option (1), so it's working as intended.
Italy tried option (2), but chickened out completely after their medical system collapsed under the strain of patients flooding every bed and corridor in their hospitals, with doctors forced to use scarce resources to treat the young while leaving the elderly to die. They are now trying option (3), locking down the entire country, like a bunch of boring mcboringstans... but it's probably too late for option (3) to stop the mass casualties for at least a couple more weeks.
Who knows? Maybe we could try option (2) ourselves, and we'll have better luck than Italy. We have absolutely no reason to believe we'll do any better, but it's a chaotic world! Who knows?
There are four countries on Earth that have successfully beaten back the exponential growth of covid. They are all still seeing cases on a daily basis, but they are not having huge numbers of cases coming in all at once and crushing their medical systems. Those countries are: South Korea, Japan, China (after the initial calamity in Hubei province), and Singapore. At the moment, it looks like they are all going to get through this relatively okay.
By a truly bizarre coincidence, all four of these countries adopted the same strategy to fight covid. That's right! They all chose option (3)! They shut down their cities, ordered widespread testing, closed the schools, and cancelled everything. And, after each country did this, cases started to fall almost immediately!
Thomas Pueyo goes into a lot of depth about how this worked in his article from earlier this week, "Why You Must Act Now." I wouldn't trust every number in his post -- some of his speculation about the "true" number of cases in any given community is a little wild, and the "model" he offers at the end is a bit high-strung as a result, but his basic point is sound: the sooner these countries imposed social distancing, the more lives were saved, and the difference was drastic.
They might be on to something!
So consider this post a prediction. If Minnesota takes no real action to combat covid, expect over a hundred thousand cases in a little over a month.
If Minnesota does take real action, then you can look at this post instead as a kind of alternate universe. Every time you miss out on something you were looking forward to doing, you can look at this post, compare my predicted case count with whatever the actual case count is that day, and thank God and your fellow Minnesotans for moving Heaven and Earth to spare us from that.
Happily, there are signs that Minnesota will indeed take appropriate measures. Several local colleges are going online-only, the Archdiocese has dispensed everyone from the requirement to attend Sunday Mass, and locals seem to be stocking up for impending social isolation appropriately.
Don't panic. This isn't the end of the world. It's not like that movie Contagion, where most of the infected died. Covid likely kills a little less than 1% of its victims, most of them elderly. Many more than that are left with permanent scars of one form or another. But we will get through this, just as humans have got through every other epidemic in history. And our forefathers didn't have Hulu. Or indoor plumbing.
Don't pretend. This isn't a run-of-the-mill flu pandemic. Stop imagining it is. It's likely the most serious health threat the human race has faced since the Spanish flu almost exactly a century ago. The fact that bodies aren't stacked like cordwood outside hospitals doesn't mean covid is a hoax; it simply means that covid hasn't taken root in your community yet. Again, look at those mass graves in Iran. (I'll be writing more about this tomorrow / later today.)
Do prepare. If you're in Minnesota, it's time to start working from home if you can. I hope you did the sensible thing and stocked up on essentials over the past few weeks; if not, there's still time. (However, please DO NOT buy masks -- health care workers need them much much much much much more than you do.) Stop going out for non-essential activities. Don't gather with large numbers of people. Take care of yourself and those around you, especially the elderly. Do your part, and we're that much closer to "bending the curve" and preventing the awful predictions in this post from coming true.