A man with one watch knows the time. A man with two watches is never quite sure.
And so it goes with methods to impute the “true” counts of official new COVID-19 cases, based on the increasingly sketchy reporting. I now have two methods for doing this — neither of which is without flaws — and both are telling me that, out of the blue, U.S. new case counts are now rising.
Let me start with Canada. Because, as you know, they celebrate Thanksgiving on the second Monday in October. (Thus efficiently guaranteeing that it’s always just a three-day weekend). My point being, I worry about the U.S. data being messed up by Thanksgiving non-reporting, putting an uptick into the current data. But that’s not a concern for Canada.
And here we are. All at once, a big lump of cases. Just this past weekend or so. After months of a steady new-case rate.
Source: Our World in Data, via Google search.
OK, with that as prologue, on to the U.S., where … no matter how I cook the numbers (old imputation or new one), there appears to be a sharp uptick in reported cases, coincident with that occurring in Canada. (Here, I’m showing my newer algorithm, because that’s what I’ve been using for the past couple of weeks.)
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 12/6/2022, 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
At some level, abrupt change is to be expected when most states report only once a week or maybe once every two weeks. All I can do is assume that the prior rate continues to hold, until the new data show up. And if there’s an upward trend, that approach to gap-filling the state data is going to show some abrupt shifts. Basically, a week or two of trend shows up in a single day.
In the Northeast — with that exceptional growth rate — both New York and New Jersey saw big upward blips in their most recent data reporting.
But I think what clinches it for me is Virginia. We’ve got pretty good data, we continue to report five days a week. Based on the reported data, and my best guess for gap-filling anything missing, we’ve gone from about 10 cases per 100K per day on Thanksgiving, to more than 17 with this most recent bit of reporting.
A final bit of confirmatory data comes from CDC, who tentatively shows a sharp uptick in new COVID-19 hospitalizations coincident with this apparent increase in officially reported cases.
Source: CDC COVID data tracker, accessed 12/6/2022
One final caveat is that all of this might just be a side-effect of the current flu epidemic, which has now reached Code Red in much of the country. I’m guessing there’s a lot more people going to see the doctor for respiratory infections right now, and that might translate into more testing. Which, a week or two later, would translate into higher COVID-19 officially-diagnosed cases.
That’s speculation.
For now, all I can say is that officially-reported cases have — out of the blue — begun to rise in Canada and the U.S. That coincides with an uptick in hospitalizations.
On balance, I think this is real, and not just an artifact of reporting.
