Thread by @dhmontgomery

So today’s #COVID19 case data in MN is astonishingly bad — so bad that my first thought is there’s a data issue. I’ve inquired with @mnhealth to see if something’s wrong, and will delete this tweet if it turns out there is.

Short story: cases up, tests down, leading to this:

@mnhealth Here you can see newly reported cases near record levels, while newly reported tests fell to a typical Tuesday low value.

I wonder if some lab only reported their positive tests today, and not their negatives?

@mnhealth There’s not a big backlog of old cases in today’s data:

@mnhealth Now, last Tuesday’s data also saw a positivity rate spike, to nearly 10%. But that was from a higher Monday base, so the magnitude of the jump was much smaller than today’s. But it’s possible there’s something new about Tuesday data (affected by weekend reporting)…

@mnhealth Alright, bad news: there is NOT a major data issue behind today’s spike in positivity rate.

That said, never put too much weight on one day of data. Focus on the rolling average. The next few days might bring continued elevated positivity rates, or a return to more normal levels

@mnhealth For example, look at the one-day spikes in positivity rate in early August and early September here. Both were scary, but were followed by sustained declines in positivity rate.

Alternately, today’s data could be a sign of a new, scary phase in MN’s outbreak.