Again, everyone's getting hung up on death numbers and mortality rates.
Guys.... these are irrelevant at the moment. Pretty much utterly and entirely irrelevant.
And that's true, even without taking into account that BBC article which points out - quite correctly IMV - that coronavirus death numbers may well be simultaneously being overcounted and undercounted, depending on the body reporting.
The only numbers that are worth watching are the number - and especially the ongoing rate of change in that number - of newly identified COVID-19 infections, and the later subset of that, the number of identified cases requiring hospitalisation.
Because that's what the objective of the imposed social distancing measures is all about - suppressing the R0 number of the infection, keeping the infection growth rate down and reducing required hospitalisations.
Succeeding in doing that will of course eventually result in less deaths - but that shift in mortality trend won't be visible for weeks. So why lhe Hell would anyone look at a data point that literally cannot help but be at least 4 weeks behind the curve in displaying anything useful or enlightening?
Currently, the US new infection growth rate is somewhere between 7 and 8% when averaged over the last few days. This means that the US is currently adding around 34,000 newly identified coronavirus cases every 24 hours. If this rate merely persists for just another 7 days, by April 15th, the US will be adding around around 50,000 new cases every 24 hours.
One hopes that Hashi's guesstimate of the infection rate then plateauing over the following 5 days, then starting to drop turns out to be accurate... because if he's wrong, then the infection numbers start to look more than concerning as one progresses throughout the rest of April.
But even even if on April 16th there was all of a sudden zero new cases, the COVID-19 death numbers would continue to climb for a whole month after that point, bearing in mind treatment time combined with reporting delays - because of the cases already in the pipeline.
Hence why, although I've made a few hopefully informed predictions, namely:-
At least 750,000 US infections by April 15th (though I admit I might have overcalled this by a day or two - it might take until the 18th to hit that figure)
At least 33,000 US deaths by the end of April (that figure I'm certain I've undercalled)
At least 100,000 US deaths by the end of 2020
At least a 7 figure number of global deaths by the end of 2020.
Those last two figures I'm fairly sure I've undercalled too, but will wait till the end of April to see how things are trending then.
As I said to Obi a few posts back, come May 1st, with the benefit not of hindsight, but of a view on what the trends are looking like then, I'll happily admit I was wrong, if that's how things look then. But if by that point, the US has over a million cases of infection (which it will) AND has a new infection growth rate of 5% or more (which it may well, though this is less certain), then I won't be apologising, because the month of May will then deliver some pretty nasty additional numbers.
Hashi Lebwohl wrote:You specifically asked me for a number of days, not a number of deaths. I answered the question as it was asked.
Entirely agreed and I wasn't intending to suggest anything else. But as per stated earlier, the question I posed was about growth curve in identified COVID-19 infections... I'd never have asked you for an irrelevant number of deaths estimate.
As above, I sincerely hope that your guesstimate that the US will be over the crest of the peak of new infections by Apr 20th is bang on accurate. Sadly from looking at the daily audited figures (and yes, just on confirmed infections), I ever more strongly suspect that you're bang on mistaken - but we'll see soon enough.