COVID-19 June 12, 2020

Today is a perfect example of how screwed up the data is and how difficult it is to analytically follow the pandemic.  The data from yesterday shows -3926 cases and -87 deaths in Puerto Rico.  Yes, those are minus signs.  As of yesterday PR decided to reclassify how they count infections and deaths.  Only confirmed cases and deaths (i.e., through testing) are now counted and presumptive and likely cases and deaths are not.  Fine.  Do it how you want.  But, they adjusted their numbers in one fell swoop on a single day.  So, what I have now is a series of daily numbers that include confirmed and presumptive cases with no way of knowing which are which.  And then, yesterday, a big adjustment.  What am I supposed to do with this?  How are those 3926 cases and 87 deaths distributed over the prior data?  Who knows.  Well, someone knows, but they aren't telling.

PR is not unique.  These types of adjustments happen periodically in many states and there is insufficient information to go back and retroactively adjust the prior daily numbers.  So, today's case numbers have a big 3926 chunk taken out that isn't reflective of what happened yesterday.  And there are 87 deaths subtracted out.  Obviously, that's going to mess up daily and 7-day trends.  But, there you have it.

PR is one of the states/territories that I've been watching as a potential hotspot.  I try my hardest not to be conspiratorial, especially these days, but it is coincidental that this adjustment is being made right when things appear to be exploding.  Other states like Texas, Florida, Arizona, California, and Utah (this is not an inclusive list) are also starting to see upward breakouts in cases.  This is unlikely to be entirely due to increased testing.