Quote (Kayeto @ May 5 2020 02:40pm)
"Reported number is question was 61k deaths as of April 30"
That's a national count. Do you think that the real national number can be both higher and lower than 61k?
I'd refer you to the CDC's excess death reporting:
https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htmif you examine the data, its clear that the excess death toll is
estimated around 61k (now 66k), but that is entirely the product of estimated numbers of unreported deaths during the lag time as it takes ~8+ weeks for deaths to be accounted. Without weighting for underreporting, the excess death toll drops from 66k to 12k, and we won't know where the truth actually shakes out for some time. There's reasonable explanations that could lead to either the the estimate being high or low, even if the deaths to the disease is kept constant. Perhaps the alertness and prompt focus on medical attention and reporting is causing deaths to be reported to the feds much sooner, in which case the estimate is too high. Maybe everyone sheltering in place and locking themselves up and cutting off social contact will lead to many unreported deaths of those living alone, in which case the estimate is too low.
What I can say for sure, is that nobody should be
sure whether that 66k figure is accurate. Its an estimate based on weighting the numbers, not reported deaths. The real national numbers can be higher or lower than it. And in some cases like when New York City was reporting 10,000 coronavirus deaths in a span with only 3,000 excess deaths, its very clear their numbers were wrong.
1
But to put numbers to it: For the week of April 11th, the CDC has 65,100 reported deaths on a threshold of 58,154 expected non-excess deaths, so 6946 excess reported deaths. But weighting for unreported deaths, they estimate 79761 total deaths, so 14661 unreported (18.4%) for a total of 21607 excess deaths reported & unreported. So the uncertainty in their arbitrary weighting estimate is the difference between a range of 7000 -> 21600 dead. How reliable is that estimate? It could be too low, or too high
This post was edited by Goomshill on May 5 2020 02:00pm