Sunday, August 16, 2020

Infection Rates

 IHME reports nationally "confirmed infections" (which I think means new) going from 78,654 on 7/20 to 67,226 on 7/30.  Curiously some days show 0.  Which I think tells you what to know about their methodology.  I thought they used to show a smoothed average for new infections.  I know they used to show number of available beds, and ICU beds as well, but their absence tells me that there is no longer any accurate data or there is not enough drama showing how far below capacity we are.

In fact, none of this data is past late July.  Why?

The Idaho news is pretty good.  7/26 had 913 new infections; 8/3 had 560.  Idaho's website shows 304 new cases ("confirmed and probable") on 8/15.  This Idaho Dept. of Public Health page graphs new cases by day with 7/15 being the peak of 601 confirmed new cases and 276 confirmed as of 8/15.  These are all date the state was notified, so a smoothed average would be more useful.  The trend is definitely down.

Again: stay masked, wash your hands, and avoid large gatherings until the "sky is falling" crowd stops screaming (likely the day after the November election).

1 comment:

  1. A "confirmed" is one confirmed by a test, like the gold standard RT-PCR that's been used since before Jan 11th. After some CDC confusion, this does not include serology tests that just check for antibodies, i.e. the person had COVID-19 in the past. There are new serology tests that look for current infections, but inherently have a higher false positive rate, an acceptable tradeoff if they're cheaper and/or faster. See also the new saliva test, don't know anything about it.

    As for hospital utilization data, based on my home state's experience, squeezing that out of them seems to be very hard. Whereas a positive current infection test result is "reportable", by law must be reported to public health authorities, like TB and other deadly communicable diseases.