Welcome back to COVID in Minnesota TODAY, an email newsletter from MPR News about understanding the latest developments in Minnesota's COVID-19 outbreak.
It's been a bit of a delay since the last newsletter, but don't worry, I've got some good stats cooked up for you!
The general picture of COVID-19 in Minnesota hasn't changed too much in the past two weeks. New infections are still in record-low territory and still getting better. Other metrics are also improving, though deaths are stubbornly high. Meanwhile Minnesota's vaccination campaign is sputtering along at a snail's pace.
I'll get into all that, and much more, below.
Three things to know:
Cases are falling to record low levels, even as vaccination rates sputter
COVID-19 tests should provide reliable data even as case counts get really low
Minnesota's third wave only saw a modest spike in death rate, but unvaccinated populations seem to have been hit much harder
Two weeks ago I reported that Minnesota was averaging "just 332 new confirmed cases per day." Fortunately, those dark days are in the rear-view mirror. Today, Minnesota is averaging 143 new cases per day.
The decline has been so steep and consistent and unprecedented that I've had to truncate most of my charts so you can see the recent changes.
And remember, those days in March and April 2020 where we last had cases at this level had extremely limited testing volume — to the point where it's not really comparable. The current average of 143 cases per day comes on nearly 13,000 tests per day. In mid-April 2020 we averaged around 1,400 tests per day, an order of magnitude lower.
As such, instead of a 10 percent positivity rate like we saw in April, today we have a minuscule 1.1 percent positivity rate.
On false positives
A lot of people have asked lately how low these positivity rates can go. Many medical tests have significant false positive rates. If that's true for COVID-19 tests then we might expect them to have a "floor" below which the positivity rate will never go.
A non-negligible false-positive rate on COVID-19 tests would also mean that the less prevalent COVID-19 is in Minnesota, the less useful tests would become. This can be a bit counterintuitive, so here's a hypothetical example:
Let's say that 1 in 1,000 Minnesotans currently are infected with COVID-19. (The actual rate of active, confirmed COVID-19 cases is about 0.2 in 1,000, or one-fifth of this made-up number for total confirmed and unconfirmed. In other words: It's not a totally outlandish number.)
Now let's say — again, making up numbers here — that if someone does have COVID-19, the test has a 95 percent chance of correctly identifying it (the false negative rate). And let's say — again, all these numbers are made up — that if you don't have COVID-19, it has a 1 percent chance of mistakenly saying you do (the false positive rate).
So all these made-up numbers were true, and your COVID-19 test came back positive, what would it mean?
Somewhat counterintuitively, if you got a positive result on this hypothetical test with a 5 percent false negative rate and a 1 percent false positive rate, the odds would be overwhelming that you didn't have COVID-19. There would only be an 8.7 percent chance you actually had COVID and a 91.3 percent chance you didn't. Even tiny false positive rates can have a huge impact when a disease is rare.
Now let's consider an alternate scenario where COVID-19 is much more common in the population. Let's use the same false positive/false negative numbers, but say that instead of 1 in 1,000 having the disease, 30 in 1,000 do. (This is a vaguely plausible figure for where things might have been back in the November wave.)
With this disease prevalence, if you tested positive, there would be a 75 percent chance you actually had COVID-19.
(You can learn more about the math behind this, called "Bayes' Theorem," here.)
Same test, wildly different results — all depending on how prevalent the disease is. So you can see why the question of the false positivity rate for COVID-19 tests matters so much.
This means a few different things. First, and most importantly, we can still trust this data even as the disease gets rarer and rarer.
Second, there's no real "floor" for COVID-19's positivity rate. It can go way lower than our current 1.1 percent. In fact, Israel is currently reporting a positivity rate around 0.1 percent, as this chart from Our World In Data shows.
Hospitalization figures are also still showing good news. Fewer people are being hospitalized with COVID-19 than basically ever before.
The death rate remains a much more ambiguous metric. Minnesota is still averaging around six or seven deaths per day. That's an improvement from where we were in May, when nine to 12 deaths per day were more common. But we haven't seen the steep, sustained declines in death stats that we've seen in cases, positivity and hospitalizations.
As I've said before, this isn't necessarily shocking, since deaths are a lagging indicator, and there are signs that the lag may be getting longer due to better treatment, differences in who's being infected, etc. Eventually deaths will drop meaningfully, given the declines we've seen in cases and hospitalizations. It just hasn't happened yet.
But a closer look at the data gives us a better explanation at what might be happening here.
Remember, more than 50 percent of Minnesotans are currently vaccinated, and this rate is highest among the most vulnerable age groups.
And if the vaccines are working at all (everything says they are), we'd expect these vaccinated people to get sick and die far less. But significant numbers of Minnesotans, including 10 percent of older people and one-third of adults 50-64, aren't vaccinated.
Overall, Minnesota's COVID-19 death rate per 100,000 people is down to about 0.1 per day. But the death rate per 100,000 unvaccinated people is 2.5 times higher.
And deaths aren't distributed evenly among all age groups. Older people are exponentially more likely to die from COVID-19 than other age groups, and middle-aged adults are much more likely to die than younger adults — though there have been recorded deaths in every age group.
Unfortunately the health department's age data isn't great. They've changed the brackets they share it in several times over the course of the pandemic. It used to come in 10-year age brackets, for example, which is great except you can't cut off at age 64, for example. Now they do share data that cuts off at age 64, but we can't compare it to the earlier data. And we still don't have case-age data for key vaccine cutoffs, like 16- to 17-year-olds.
But even with those limitations, we can get some interesting insights. Here is a chart showing the per-capita COVID-19 death rate for people 50-64 and 65-plus, from fall 2020 through today. As before, I've broken out the rate per 100,000 people, and the rate per 100,000 unvaccinated people.
The most striking thing is the left graph for middle-aged adults 50-64. If you're in this group and unvaccinated, the relatively mild wave this spring was actually deadlier than the fall wave, and even after the third wave has crested and receded, this group is still
dying at rates comparable to last fall. It's just that 70 percent of this age group is vaccinated, so the overall figures are pretty low.
It's a similar story with older people, where the unvaccinated 10 percent is having elevated death rates about half the fall peak.
All this suggests that the spring wave might have been way worse if Minnesota's vaccination campaign hadn't been as advanced as it was. (Remember the lag from infection to deaths — many of the people dying this month got sick back in April or May.)
One final point of clarification: These graphs aren't directly measuring the risk for unvaccinated people. Obviously if you take the same number of deaths and divide by a smaller population, you're going to get a larger number. But I think they approximate it as closely as population-level statistics can. Minnesota has had very few fully vaccinated people die of COVID-19, just 44 of well over 1,000 deaths since vaccinations began.
Below, I'll explore a few more stats about Minnesota's COVID-19 outbreak. If you want to learn more, I also recommend these other articles about the COVID-19 outbreak:
Minneapolis metro over 70 percent vaccination target, other areas lag
Minnesota's vaccination rate has been slowing to an utter crawl. Over the past week, we've averaged fewer than 5,000 new first doses per day. Back in early June, when I last wrote you, we were down to 10,000 per day, and it seemed terrible. But it wasn't a floor.
Right now, 10,000 doses per day would be pretty good! Minnesota is close enough to its target of 70 percent of adults vaccinated — we're currently at 65 percent of adults — that 10,000 per day would get us there in a few weeks.
At the current rate of less than 5,000 new adult first vaccinations per day (remember, the chart above was all vaccinations, this is just those 16+), we're looking at a month and a half to go.
The vaccination rate is petering out just about everywhere — all age groups, all regions of the state. But not every region of the state is stalling at the same rate.
Right now there are seven counties with more than 70 percent of adults vaccinated: Cook, Olmsted, Hennepin, Ramsey, Washington, Dakota and Carver. Another three counties are on the verge: Carlton, Houston and Scott.
Vast swathes of northeastern and southeastern Minnesota are above 60 percent. But then most of western Minnesota is below 60 percent vaccinated, and a big chunk of counties north of the Twin Cities metro is below 50 percent.
The geography of COVID-19 right now in Minnesota
I previously shared an interesting map based on a New York Times algorithm for classifying the COVID-19 infection risk for unvaccinated people in a county. You can read my full writeup
from earlier in June, but just know that it's based on per-capita case rates, except for very small counties where it uses the raw number since a few cases in a small population can skew rate stats.
Here's what that chart looked like as of June 1:
And here's what it looks like today:
As you can see, case rates are drastically improving just about everywhere, except in the north metro and a big block of northern Minnesota.
There are actually a dozen counties in western Minnesota, mostly quite small, that haven't recorded a single new COVID-19 case in the last two weeks.
Nobles just has a 60 percent vaccination rate, but all those prior infections confer immunity, too. It took more than a year, and 50 deaths, but Nobles might be finally seeing the silver lining from being arguably the state's hardest-hit county: herd immunity.
What's happening next?
Hopefully I won't make you wait another two weeks for the next newsletter. I'm working on some interesting stats about "breakthrough cases," or infections among previously vaccinated people.
Thanks for subscribing to this newsletter and for all of you who've sent me your feedback. As always, if you like this newsletter, please share it and the sign-up link with others who you think might enjoy it.