COVID-19 (2): Turns out it's going to be pretty bad actually

What kind of ventilator uses a sprinkler head???

:sweat_smile: at least thatā€™s how I read it the first time. :blush:

Iā€™ll take more than 60,400 confirmed deaths by 6/4

Lol at 8/4

https://mobile.twitter.com/RawStory/status/1248458860613926917

Why do you think we are going to eradicate this and itā€™ll just go away and have no other peaks? You think we will stay at home for the next 18-24 months waiting for a vaccine and not open back up? You think NYC being almost half of the US deaths when itā€™s hitting a peak and the rest are just ramping up means the others will magically die and never ramp up?

80-100k are summer numbers, not final numbers

1 Like

That would be great; fingers crossed

Doubting itā€™ll work well for something as severe as covid tho, but if it does itā€™d be great.

This makes it sound like NYC is out of resources for all intents and purposes and just stretching them thin enough to pretend people arenā€™t being sent home to die.

People get sent home, they get 20 minutes of paramedics (or worse, EMTs) treating them, theyā€™re pronounced at home. People get to the hospital needing a vent and have to wait for someone on a vent to die so they can get one. Their family isnā€™t with them so we arenā€™t hearing about these things, but theyā€™re happening.

2 Likes

That is rough. Plus the EMTs have to call time of death now, doctors are not brought online to announce time of death anymore.

I keep hearing all this stuff that we are getting past the worst but reality keeps pointing in a different direction.

I trusted the numbers going up, so going to have to have faith as they level off but still seems we are a long way off from good times in New York and a good portion of the country is much further behind the New York timeline.

Ok Iā€™ve been doing some maths. Having fun with the new xlookup function in excel.

For all states/territories on covidtracking, I took the raw 7 day exponents (Ln(today/last week))/7. Not slopes. Should minimize the variation based on reporting days over weekends. Monday/Monday, etc.

I then projected 7 days forward. The net average for the whole US works out pretty close to my slopes average exponents. Currently about .14 (5day doubling total deaths)

7 days forward projects to about 45,000 total From 17,000 now. Weā€™d need to see an average of about 4,000 per day. It is massively dominated by NY (50%), Then NJ (14%), MI (9%)

HOPEFULLY NY is leveling off at about 6,000 a week. The exponent is moving strongly in the good direction (expected when the rate goes flat). Too tired now to factor this in today.

Eventually I may Try to build this out more based on positive tests

Speaking of that I looked at NY only.

I took death rate (total deaths/total cases). I then did it at 1,2, ā€¦ 20 day offsets. 7 would be total deaths today/total cases a week ago.

I then did the relative standard deviation Encompassing the last several days it looks like the RSD is lowest when the offset is 9 days.

It works out to 9.3% +/- 0.15%. (1.6% RSD)
That is shocking. This ainā€™t the regular flu even if we are undercounting cases more than deaths.

If calc from todayā€™s 160,000 NY cases we get 14880 total deaths on 4/18. If we are flat then we could expect to see 868/day on average from now until then. 779 and 799 the last two days. Seems like this will be a better predictor than the exponent once we see flattening.

I did a daily/daily calc for a 9 day offset and the average is about 9%. RSD is higher but expected for daily numbers with weekend biases etc.

9 days from positive test to death seems to make a lot of sense. There seems to be a few day lag between exposure and a positive test. I think I saw a factor of 14 days from exposure to death on average.

*standard possible biases in testing and counting deaths

7 Likes

And then I read the NY paramedic deal and it just slams that the ā€œfunā€ math excel Iā€™m doing is death of people. Dying alone hooked up to a vent or dying at home just because thatā€™s where NY is at. For the state, the top 10 causes add up to 315/day (2017). 2.5 fold that just from covid and that is likely really 4-5 fold.

Covid is the #1 killer in the country by daily rate.

Rachel said someone every 47 seconds today in the whole country.

Nah, I had it way back in November. I was with the guy who started it when I dared him to eat a raw bat. Didnā€™t think the crazy bastard would actually do it.

6 Likes

She canā€™t come to your house

If the sister isnā€™t happy get a hotel room

1 Like

want to feel old? this was one month ago:

6 Likes

Joe or surrogate (Warren or Porter would be perfect) needs to do a daily 15 minute talk with a 15-30 minute QA with a couple of real unleashed experts.

In one of them they could have a timer ding with a counter for every damn death during the press conference.

Let Joe MC. Give the progressive wing lots of love. States will carry it live when itā€™s their reps turn.

1 Like

https://mobile.twitter.com/brhodes/status/1248454471388020737

somebody has to get this img right? itā€™s not the deepest cut

@RiskyFlush is your huckleberry

It says the total rate of infection (active and already recovered ) is 15%

1 Like

My Blue Heaven?

2 Likes

But also the conversation happening.

Youā€™ve got me there. Not sure I ever saw it. Please explain in your best Roger Ebert voice and I will be forever in yer debt