Can’t speak for anyone else but in my area nursing home residents with covid symptoms are not getting shipped out or tested until it’s vent time. Our call volume has actually been down.
2 Likes
Trump will send ventilators as needed.
1 Like
Narrator: They did not block the GOP.
1 Like
1600 more DOA calls (400% increase) compared to last year by NYC paramedics.
https://mobile.twitter.com/justinhendrix/status/1247979994757959683
That is 150 deaths a day in New York.
I’m guessing that the 22k cases showing up today for Nevada on worldometers is a bug?
1 Like
Electronic forms. People getting text messages when they drive across the border using Fed emergency sms system. IDK about this action but gov has been somewhat reasonable if not entirely consistent.
I wonder if some people are also choosing to die with family instead of alone in a hospital
4 Likes
That’s a good point. There have been tons of stories about people having to hear their loved one’s last words over a FaceTime call. There was a story in the Times last night about a woman who got sent to the hospital and died and it took the family a week to even find out what hospital she ended up at.
2 Likes
On my moving log slope calc there has been a slow down in the death numbers. Doubling time in the 4-5 day range. Looks like a 10-12 day lag from when your total case number rise slowed. I am worried that the numbers are going to shift to states with worse testing/reporting than New York.
10,000 nationwide in the last week. Annualized puts Covid-19 in a close 3rd behind heart disease and cancer.
Jeannine Piro must be the most stupid person on television - even worse than Trump: „The flu would be deadly if we didn’t have a vaccine.“
Even people watching FOX news regularly must be able to detect the problem with that sentence.
2 Likes
Narrator needed.
3 Likes
Out of curiosity, I did a little digging into the guts of the IHME model, which you can find here: https://www.medrxiv.org/content/10.1101/2020.03.27.20043752v1.full.pdf
It turns out all they are doing is estimating a highly parameterized cumulative distribution function. There have probably been some updates to the methodology compared to what is described in this early version of the paper, but that I think that is still probably mostly what is going on.
In a nutshell, I don’t think this is a great approach to this problem. I might elaborate more later.
I just went outside. It’s weird out there man.
I may need to start saving my piss in jars.
So basically it’s assuming a function (Polynomial?) for the curve shape?
It’s not based on first principles such as estimated transmission rate, lag to hospitalizations, lag to deaths, etc?
As far as shape, looking at China Italy and Spain, looks like there is a leveling off for about 2 weeks followed by a decline that is more gradual than the original increase. Not sure why the IHME curves are so sharp and more or less bell shaped.
With the US having so many localized outcomes, I expect to see stair steps for the next 6-8 weeks assuming NY is at its plateau.
I also expect our effective transmission to be a bit higher than Italy and Spain (guessing we aren’t as tight as them) and we will have covidiots (trump and R-govs) that will push for relaxation too soon and secondary rises.
So our plateaus May be longer and the local declines may be more gradual.
2 Likes
She is peak Fox News. I honestly don’t understand how anyone isn’t watching her to laugh, but there are true believers for any subject.
I would however pay many thousands of dollars to see Judge Jeannine debate Janeane Garofalo
Pandemic thoughts: husbands and wives shouldn’t work in the same office. Especially if that office is their child’s preschool.
8 Likes
essentially they are fitting a normal distribution (bell curve) to the pattern of daily deaths. this is why it is predicting a sharp post-peak fall in the death rate for states that have had a sharp rise in death rate. it’s also why the model predicts 0 deaths after a few months.
they are trying to capture some information on the impact of social distancing, which they use as part of the information to set different parameters for each state.
but i think this approach is way too rigid. i would be much more interested in a first-principles type model (e.g., the SIR model Compartmental models in epidemiology - Wikipedia).