Where Are Minnesota's Tests?

Some very odd (but potentially relieving?) information in Friday's Minnesota covid presser.

The thing that worries me most about Minnesota right now, the thing that keeps me up at night, the thing that makes me confused and frustrated and worried that our stay-at-home order will have to last longer than in most U.S. states, is our testing positivity rate.

It's generally agreed, based on the countries where they've got covid under control (South Korea, New Zealand, Australia, Taiwan), that you are doing enough testing to accurately track the epidemic if your positivity rate (the number of positives divided by the total number of tests given) is 2% or lower.

As of Thursday, the median U.S. state or territory's average positivity rate over the past seven days was 11%. This high number indicates that, as a country, we are still flying mostly blind and have zero ability to get ahead of covid-19 with testing, tracing, and isolation. However, the number has been falling steadily for days, as national testing rises.

Also as of Thursday, Minnesota's average positivity rate over the past seven days was 17% -- and steadily rising. Our daily testing was supposed to hit 5,000/day a week ago (on a steep path towards 30,000), but it has yet to do so, and has instead spent nearly two weeks fluctuating between 3,000 and 4,000. This indicates that we are not only doing far worse than we need to be, but that our control over the epidemic is actually slipping... this despite a set of stay-at-home orders that are buying us limited time to get things under control. We're also one of the few states right now where, given our official case count, the virus still seems to be growing.

Let's be clear: once the stay-at-home orders end (and they will have to end within a couple months, if not weeks), we either have the ability to test, trace, isolate, and suppress the virus... or we don't. If we do, we can get through this without it being too much worse than a real bad flu. If not, brace for impact, because thousands of Minnesotans are gonna die. We don't get a do-over. We need our positivity rate at 2% or lower YESTERDAY. And we need to start seeing daily case counts and death counts fall very soon if we plan to get it to a controllable point by May 30th (I don't see how May 18th is viable at this point).

So that's context for two tidbits from this hourlong Friday briefing that really caught my attention:

(1) Commissioner Jan Malcolm reports early in the call that Minnesota just found out this week that some testing sites are only reporting positive tests, not total tests. This means that the positivity rates for those sites is being recorded as 100% -- but that's not actually accurate. Minnesota is now trying to correct that data. We don't know how widespread this is, and whether it will have a meaningful impact on the size or direction of our positivity rate, and it's... surprising... that this reporting error was allowed to persist up until now. But HOPEFULLY, once the data is corrected (SOON PLEASE), we will see that we've actually been conducting more tests than we realized.

[UPDATE: Commissioner Malcolm clarified later in the day that this only happens occasionally, and that, on May 7th, only one lab did this. That's good news for the competence of Minnesota's data scientists; bad news for all of us who were hoping it could explain away our high positivity rate.]

(2) Even more interestingly, Commissioner Malcolm and Governor Walz answer a question by stating that Minnesota, right now, today, has capacity to test 10,000 people per day. That capacity is not being used. According to Commissioner Malcolm, people are literally not showing up at testing sites to get tested, possibly because of communications months ago that we don't have enough tests and that mild cases shouldn't get tested.

This has changed! MILD CASES SHOULD GET TESTED NOW. If you show symptoms of covid-19, get thee to a testing site!

But apparently our own Minnesotan politeness -- the same dumb obsequiousness that makes us a joy to grocery-shop with and a nightmare to zipper-merge with -- is keeping us from using our full testing capacity, getting a true handle on the disease, and getting out ahead of the rest of the country as we should. (That, or the Governor is not telling us the full story. Always a possibility with government officials.)

Walz and Malcolm vowed to communicate this better. They'd better be quick, because this has been turning into a farrago for Minnesota even as North Dakota whups our butts on testing and cases and deaths.

In Other News: Minnesota has received its first batch of remedisivir, which I still can't spell after two months. If this thing is capable of cutting covid's fatality rate by even half, I'll feel very comfortable about our position, tests or no tests. One of the reasons we locked everything down was to buy time to develop treatments, and now those treatments are finally starting to arrive.

Q&A: Collected from readers over the past few days.

I heard that Minnesota will peak late because we did a good job slowing the spread. Is that true?

I don't see why this should be true. Peaks happen in one of two ways: (a) achievement of herd immunity forces the disease into decline, or (b) human interventions like vaccination or decisions to stay at home (legally mandated or not) force the disease into decline.

No state has reached herd immunity, not even New York City (where it seems about 20% have been infected; herd immunity is expected after around 75% have been infected).

So all states that have passed their peaks (their first peaks, that is) have done so by staying at home and other measures, like washing their hands and wearing masks. (They will immediately start toward their second peaks when they lift those orders, unless they have some suppression in place OR people continue voluntarily staying home.)

It is very frustrating (and confusing!) that our very similar measures have not had this same effect so far. Hopefully it's just a bizarre fluke in our relatively weak testing... but, every day that our positivity rate stays high, that becomes less likely.

Are we seeing Minnesota's positivity rate go up because we have more tests now and are detecting more cases we weren't detecting before?

We are doing more tests, so we should be detecting more cases -- and we are. That, in itself, is not alarming. It should actually be reassuring. The absolute case numbers were bound to go up once we got our act together with testing.

But, if our testing is actually catching up to the virus, our positivity rate should be going down. We should be detecting more cases overall, yes, but we should also be detecting far more negative results as we test more and more marginal and questionable cases. (That's how you discover all the real cases -- by testing everyone who might have it.) We're not seeing that. The ratio of positive results to negative results has been rising, which indicates that even our increased testing hasn't kept pace with the spread of the disease.

Are we really doing worse than most other states in terms of testing?

We're not abysmal. There are other states with much lower per-capita averages, and (if you pull out a spreadsheet and do the math) lots of states with higher positivity rates over the past 7 days.

We in Minnesota are significantly worse than average, both on per capita testing and on positivity rate. To everyone who shares my serene confidence in the competence of Minnesota's bipartisan government and in the character of its hardy people -- confidence that has been consistently rewarded for decades -- this is inexplicable.

Remember how everyone took a few days to bash Florida a few weeks ago? I did a bit of this on Facebook, too. Well, they seem to be kicking our butts on both measures.