Saturday, November 28, 2020

C19 Update with Apparent Mortality

Looking at my C19 cases/deaths plots today I wondered how many cases there were for each reported death. Using the 7-day average for each, this would be a kind of apparent mortality metric. I say apparent because the number of cases is known to be undercounted by up to a factor of 10. Here is the plot. Possible interpretation of the US curve: Feb is initial spike (off scale, max 16%); mid Mar to early Apr deaths ramped up faster than testing could be expanded; early Apr to Jul more widespread testing and better care methods brought the values down; since Jul slightly improving care and testing. If the health care system gets swamped, I would expect the apparent mortality to increase as quality of care erodes.