« prev   random   next »
▲
1
1
▼

Projected 100K to 200K fatalities in US by August still likely

By ThreeBays follow ThreeBays   2020 Apr 18, 11:08am 1,042 views   39 comments   watch   nsfw   quote   share    

A few weeks ago the White House was sharing projections of 100K to 200K deaths in the 1st wave. Since then, projections by IHME model [1] revised down to around 60K (with an error range of 34K to 140K). The conclusion almost everyone has made is that social distancing worked better than expected, and so the death count would land well short of the earlier 100K+ projections.

Unfortunately, if you read details for the IHME model they are making certain assumptions about social distancing measures. Firstly, they are using Wuhan as the model for social distancing. Wuhan as we know had a more draconian lockdown and fever clinics that we don't have here. Secondly, the IHME model assumes that all States implement the strictest lock-downs within 1 week if they haven't already. Many states haven't implemented the strictest measures for example Texas. Thirdly, the model doesn't account for any state starting to re-open it's economy.

From daily trends (cases & deaths) it's clear that before mitigations started, the spread in most countries doubled every 3 days. That's a 10x increase every 10 days.

After mitigation, China deaths approximately halved every 10 days. This is exactly what IHME is projecting for the US. But US and other Western countries are not China. Italy & France are halving every 50 days. Spain is halving every 30 days. We already have estimates that New York is on a similar slope to Italy. We've only recently started asking people to wear masks and face coverings, so fingers crossed we improve on this otherwise I think we're on track for over 160K deaths through August.

Since you, me, everyone wants the economy to re-open the number is likely to be higher unless governments implement very strict isolation of vulnerable people - for example in the UK those over 70 or with certain high risk conditions have to lock in and get food & medicine delivered.




[1] IHME model https://covid19.healthdata.org/united-states-of-america
1   Y   ignore (3)   2020 Apr 18, 12:16pm     ↓ dislike (0)   quote   flag      

The only thing that chart shows is that the usa is paying a heavy price for its fructose addictions...
2   ThreeBays   ignore (4)   2020 Apr 20, 9:49pm     ↓ dislike (0)   quote   flag      

Nice to see all the usual suspects covering their ears and singing "la la la" on this one.

For anyone serious, here's some good reading regarding the flaws of the IHME model: https://statmodeling.stat.columbia.edu/2020/03/29/the-second-derivative-of-the-time-trend-on-the-log-scale/

"Fourth, yeah, what Dorman says: you can’t take the model for the asymptotic limit seriously. For example, in that methodological appendix, they say that they use the probit (“ERF”) rather than the logit curve because the probit fits the data better. That’s fine, but there’s no reason to think that the functional form at the beginning of the spread of a disease will match the functional form (or, for that matter, the parameters of the curve) at later stages. It really is the tail wagging the dog."

TLDR: The model can match the inflection point of the curve, but has little predictive power of what happens after the top. It's assumption is a drop symmetric to the rise, however as trends from Italy, Spain, France, Iran and early post-peak trend for US, the curve does not drop symmetrically down but rather extends to the right.
3   Ceffer   ignore (6)   2020 Apr 20, 10:36pm     ↓ dislike (0)   quote   flag      

Thank you for being a one stop shopping center for all the sources of information that are never to be trusted again. Saves a lot of effort.
4   HeadSet   ignore (2)   2020 Apr 21, 7:56am     ↓ dislike (0)   quote   flag      

Note that Sweden is not on the graph....
5   WookieMan   ignore (7)   2020 Apr 21, 7:59am     ↓ dislike (0)   quote   flag      

We're all dead. Not sure how I'm even typing. Weird.
6   Al_Sharpton_for_President   ignore (6)   2020 Apr 21, 7:59am     ↓ dislike (0)   quote   flag      

What do the numbers look like with the assisted living and nursing home fatalities removed?
7   WookieMan   ignore (7)   2020 Apr 21, 8:05am     ↓ dislike (0)   quote   flag      

willywonka says
What do the numbers look like with the assisted living and nursing home fatalities removed?

Like nothing, but no one wants to admit that. Death is bad, but if you wanna take the positive out of this, the average monthly SS check is ~$1,500 and can go up to $3k. 100k dead would be $150M a month save in SS minimum. That's not including medical which is much higher than the SS cost. Every nursing home death is likely saving $3-4k/mo per person dead. And remember, these are people that weren't producing anything.

I know my position is harsh on this and those with sensitive feels will dislike it. But we need to open up, and if we burn off $3-4k/mo in obligations per death, you can't really argue that's a bad thing.
8   Al_Sharpton_for_President   ignore (6)   2020 Apr 21, 8:14am     ↓ dislike (0)   quote   flag      

WookieMan says
willywonka says
What do the numbers look like with the assisted living and nursing home fatalities removed?

Like nothing, but no one wants to admit that. Death is bad, but if you wanna take the positive out of this, the average monthly SS check is ~$1,500 and can go up to $3k. 100k dead would be $150M a month save in SS minimum. That's not including medical which is much higher than the SS cost. Every nursing home death is likely saving $3-4k/mo per person dead. And remember, these are people that weren't producing anything.

I know my position is harsh on this and those with sensitive feels will dislike it. But we need to open up, and if we burn off $3-4k/mo in obligations per death, you can't really argue that's a bad thing.
It's a new virus, and folks with shit immune systems in bug incubators are going to get picked off. It might be worse than the flu, in that case, as they lack pre-existing immunity and can't develop an adequate immune response before they die, if at all. There will be occasional fatalities in the healthy population, just as with the flu, although COVID-19 is not affecting young kids as much as the flu.

You can't have MD's directing economic policy, it's insane.
9   ThreeBays   ignore (4)   2020 Apr 21, 9:34am     ↓ dislike (0)   quote   flag      

WookieMan says
We're all dead. Not sure how I'm even typing. Weird.


After you :-).

Seems like "it doesn't matter, we all die anyway" is really "others should just die, so that I can live".
10   ThreeBays   ignore (4)   2020 Apr 21, 9:41am     ↓ dislike (0)   quote   flag      

HeadSet says
Note that Sweden is not on the graph....


And what? You must not be paying attention. Sweden's curve remains exponentially up without sign of peaking. Their strategy has slowed but not stopped their epidemic.

11   Al_Sharpton_for_President   ignore (6)   2020 Apr 21, 9:57am     ↓ dislike (0)   quote   flag      

ThreeBays says
Sweden's curve remains exponentially up without sign of peaking.
You are merely describing a rate, and not the outcome.

It is possible that the sick elderly die off quickly, as suggested by Professor Johan Giesecke, and that we wind up with the same number of deaths overall, just takes longer to get there in the lockdown countries. And herd immunity develops more slowly under a lockdown scenario as well.
12   HeadSet   ignore (2)   2020 Apr 21, 10:07am     ↓ dislike (0)   quote   flag      




Seems like what was needed was to leave the economy open, but protect the elderly.
13   Y   ignore (3)   2020 Apr 21, 10:35am     ↓ dislike (0)   quote   flag      

Looks like it pays to be 100...
HeadSet says



Seems like what was needed was to leave the economy open, but protect the elderly.
14   Onvacation   ignore (6)   2020 Apr 21, 10:54am     ↓ dislike (0)   quote   flag      

ThreeBays says
everyone wants the economy to re-open the number is likely to be higher unless governments implement very strict isolation of vulnerable people - for example in the UK those over 70 or with certain high risk conditions have to lock in and get food & medicine delivered.


Why did we not do that from the beginning? Instead we have locked down perfectly healthy people with very little risk of death from this virus. The purpose was to flatten the curve; we smashed it.

Time to get back to work.
15   RC2006   ignore (2)   2020 Apr 21, 11:08am     ↓ dislike (0)   quote   flag      

Worse case a bit more deaths than bad flu season for the sickly and old. What a clusterfuck the damage that has been done but shutting shit down instead of just taking precautions. Thanks again for fucking the younger generations, and anyone that is productive and needs to work for a living.
16   MisdemeanorRebel   ignore (3)   2020 Apr 21, 11:17am     ↓ dislike (0)   quote   flag      

Remember folks: Primary voting in the Wisconsin Election will not be delayed, it's too risky to our Society.

But shutting down the economy as long as possible simply must be done.

Also:

17   FuckTheMainstreamMedia   ignore (7)   2020 Apr 21, 11:42am     ↓ dislike (0)   quote   flag      

Actually I think it depends on how well people with underlying health conditions stay the fuck at home

18   Al_Sharpton_for_President   ignore (6)   2020 Apr 21, 12:12pm     ↓ dislike (0)   quote   flag      

Yes, but no one wants to take care of granny at home, so they dump her in a bug incubator and demand we all pay the price.
19   ThreeBays   ignore (4)   2020 Apr 21, 12:18pm     ↓ dislike (0)   quote   flag      

Onvacation says
Why did we not do that from the beginning? Instead we have locked down perfectly healthy people with very little risk of death from this virus. The purpose was to flatten the curve; we smashed it.


1) It was done so as not to overwhelm hospitals, since nobody knew where the curve will actually peak. With less isolation we would have peaked higher. Recall it's not only the base fatality rate that's the issue, but also the # of people that need hospitalization or they'll die as well.

2) Maximal suppression was used to reduce the number of cases so that testing & contact-tracing can be viable.

Testing & contact-tracing is likely the game changer here. The R spread is still pretty high (0.9) with the lockdown in NY. Reopening the economy will quickly push it over 1.0 and restart the epidemic. If contact tracing can identify even 25% of cases, that's enoiugh to let us get to R 1.3 and still have a dropping curve.

Hand washing, masks, contact tracing, isolate elderly - is likely enough to go back to a near normal economy.
20   ThreeBays   ignore (4)   2020 Apr 21, 12:32pm     ↓ dislike (0)   quote   flag      

The elderly are getting the bug even with the full isolation we have. Elderly care workers need hazmat suits if the outdoors is teeming with virus.
21   Rb6d   ignore (0)   2020 Apr 21, 12:38pm     ↓ dislike (0)   quote   flag      

ThreeBays says
Elderly care workers need hazmat suits if the outdoors is teeming with virus.

Or they should be closeted inside retirement homes w/extra pay during the pandemic.
22   FuckCCP89   ignore (5)   2020 Apr 21, 12:39pm     ↓ dislike (1)   quote   flag      

ThreeBays says
The elderly are getting the bug even with the full isolation we have.


How?
23   ThreeBays   ignore (4)   2020 Apr 21, 2:55pm     ↓ dislike (0)   quote   flag      

TEOTWAWKI says
How?


People who care for them or live with them bringing it probably.
24   FuckTheMainstreamMedia   ignore (7)   2020 Apr 21, 3:16pm     ↓ dislike (0)   quote   flag      

ThreeBays says
TEOTWAWKI says
How?


People who care for them or live with them bringing it probably.


Wookie is right. Old people are stupidly stubborn as hell.
25   NDrLoR   ignore (1)   2020 Apr 21, 3:34pm     ↓ dislike (0)   quote   flag      

ThreeBays says
The elderly
are more likely to be sick in the first place and by actuarial averages more likely to die than young people. It's always been that way.
26   WookieMan   ignore (7)   2020 Apr 21, 8:52pm     ↓ dislike (0)   quote   flag      

NDrLoR says
ThreeBays says
The elderly
are more likely to be sick in the first place and by actuarial averages more likely to die than young people. It's always been that way.

That and this....

FuckTheMainstreamMedia says
Old people are stupidly stubborn as hell

This isn't hyperbole. I've legit seen electric cart oldies with an oxygen tank in the basket rolling around the grocery store in the last 7 days. This is what we're shutting down for? The one fuck that drags it back to a nursing home or retirement community? Wipes out 70-80% of them in a week. This is where we're at.

I'd just wish some people would admit they just want to politicize this instead of lying out their asses. Crazy how so many people are such pussies because one man won an election. Let's just fucking politicize this 1,000% at this point for the losers that have their dick tucked between their legs.
27   Reality   ignore (8)   2020 Apr 22, 2:45pm     ↓ dislike (0)   quote   flag      

ThreeBays says
1) It was done so as not to overwhelm hospitals, since nobody knew where the curve will actually peak. With less isolation we would have peaked higher. Recall it's not only the base fatality rate that's the issue, but also the # of people that need hospitalization or they'll die as well.

2) Maximal suppression was used to reduce the number of cases so that testing & contact-tracing can be viable.

Testing & contact-tracing is likely the game changer here. The R spread is still pretty high (0.9) with the lockdown in NY. Reopening the economy will quickly push it over 1.0 and restart the epidemic. If contact tracing can identify even 25% of cases, that's enoiugh to let us get to R 1.3 and still have a dropping curve.


Those over-simplified calculations make two erroneous assumptions:

1. The virus is genetically stable and not mutating (which is not the case for an RNA virus, which the Corona Virus is);

2. The entire population is naive (in epidemiology, it means people having never been exposed to the virus or the immune system has entirely forgotten the virus), and will become naive again the moment after being tested positive (i.e. the virus never triggers any immunal response or memory). Otherwise, the R number quickly drops towards zero as the entire population gets exposed to it quickly!

BTW, the focus on making R less than 1 through lock-down is a little like the old Stalinistic saying: one death is tragedy, but a million deaths is only a statistic. R<1 doesn't mean individuals don't get infected or get severely ill or even dying; it just means the total number of getting sick at the same time is lower, which is a meaningless statistic in terms of how many people die from it eventually . . . and outright dangerous when facing a rapidly mutating virus if the time line is stretched out!

As mutations take place, if the society is locked down whereas the hospitals are not, the most deadly mutations getting higher probability of spreading (getting extra opportunity to spread when they land the human in hospitals then spread to other people there) than the less deadly mutations (that leave people healthy therefore locked down), the lock-down of society prevents people from early exposure to small quantities of new mutations and instead force them to face more deadly mutations for the first time at high viral load at hospitals or at homes with hospital workers (i.e. for an entire night instead of short casual exposure in normal social settings). That makes the lock-down policy (while not locking down hospitals) a homicidal policy.
28   ThreeBays   ignore (4)   2020 May 5, 11:01pm     ↓ dislike (0)   quote   flag      

It's been a couple of weeks. US graphs have been moving sideways as predicted. Some states like NY have peaked and come down, while some like Florida are moving sideways. California seemed to dip and then move sideways. Illinois and Texas and and many states are still moving up. This means that infection is still highly prevalent as states try to open up their economy.

We crossed the 60K predictions, pushing on 70K. The White House's favored IHME model has completely revamped their models, now projecting 134K fatalities by end of August. Seems optimistic. https://covid19.healthdata.org/united-states-of-america

I've been using this model https://covid19-projections.com/about/ more since it seems to have had the closest predictions so far. It's currently forecasting 156K fatalities by end of July.


29   mell   ignore (6)   2020 May 5, 11:10pm     ↓ dislike (0)   quote   flag      

Daily new cases have been trending down pretty much every day, so have deaths with exceptions like today. It's normal to see a rise in states that have not been affected much prior, doesn't necessarily mean the lockdown or lack thereof is responsible for it. In each area eventually the virus will follow the same curve and those hardest hit are those to come down fastest as well. As long as hospitals aren't overloaded states should be open for business, the daily case rate is largely irrelevant at some point where enough have been exposed. Remember exposed doesn't mean actively infected so the exposure is largely much higher and will eventually provide herd immunity.
30   ThreeBays   ignore (4)   2020 May 5, 11:17pm     ↓ dislike (0)   quote   flag      

mell says
It's normal to see a rise in states that have not been affected much prior, doesn't necessarily mean the lockdown or lack thereof is responsible for it.


The states that are coming down are due to a combination of the shutdowns plus building a % of immunity. Most states haven't built a significant % of immunity so that's why they're still going up. Loosening shutdowns will make infections rise, but it'll be about 2 weeks before this is visible in cases and almost 3 weeks until a change in fatalities.
31   ThreeBays   ignore (4)   2020 Jul 31, 7:30pm     ↓ dislike (0)   quote   flag      

Was I right?
32   MisdemeanorRebel   ignore (3)   2020 Jul 31, 7:47pm     ↓ dislike (0)   quote   flag      

Fuck this thing already. Reopen everything.
33   PeopleUnited   ignore (1)   2020 Jul 31, 8:18pm     ↓ dislike (0)   quote   flag      

ThreeBays says
Was I right?


No way to know for certain, too much politics and not enough objective analysis. “Official numbers” are riddled with doubts of conflict of interest on so many levels.

Plus, how many lives could have been saved with early use of Hydroxychloroquine/zinc cocktails (meaning give them at first sign of infection) and steroids like dexamethasone?
34   theoakman   ignore (0)   2020 Jul 31, 8:42pm     ↓ dislike (0)   quote   flag      

You neglected to predict massive demonstrations where hundred of thousands of people held hands and spit all over each other.
35   mell   ignore (6)   2020 Jul 31, 9:09pm     ↓ dislike (0)   quote   flag      

ThreeBays says
mell says
It's normal to see a rise in states that have not been affected much prior, doesn't necessarily mean the lockdown or lack thereof is responsible for it.


The states that are coming down are due to a combination of the shutdowns plus building a % of immunity. Most states haven't built a significant % of immunity so that's why they're still going up. Loosening shutdowns will make infections rise, but it'll be about 2 weeks before this is visible in cases and almost 3 weeks until a change in fatalities.


There were only a few days with deaths over 1k so far since the first wave and even those are less than half of the peak deaths during the first wave, despite much "higher" daily cases (testing bias). Obviously we are testing the crap out of it with some counties already having tested 25-30% of the population. At this point it's meaningless and the cat has been our of the bag since Jan/Feb already. They should open up everything with hygiene protocols/recommendation (could be left up to the owner) for indoors and larger outdoor gatherings. The lockdowns don't have much effect on Covid and the devastating long term economic consequences and lives lost due to suicide, depression, drug use and not being able or afraid to see your MD are far worse than the Covid deaths of mostly very old and immunocompromised people. Plus in early stages its very treatable with HCQ et al but people are being sacrificed for leftoid politics and TDS.
36   MrEd   ignore (3)   2020 Jul 31, 9:45pm     ↓ dislike (0)   quote   flag      

We are all fucked because we live in a free country.
Communist regimes are the lucky ones with covid 19, where every deterrence is mandatory.
37   mell   ignore (6)   2020 Jul 31, 10:18pm     ↓ dislike (0)   quote   flag      

MrEd says
We are all fucked because we live in a free country.
Communist regimes are the lucky ones with covid 19, where every deterrence is mandatory.


Actually you may be right in this specific case as all Western countries pretty much observe the same curves, proving that non military lockdowns don't help much. However even if you believe that China's numbers aren't royally fudged, the general life expectancy in communist countries is lower than in the Western world and Covid won't even make a dent in the numbers. So you not only live better but also longer in non communist countries on average, Covid or not. And in the end military lockdowns only delay infections and deaths at a much higher price.
38   Patrick   ignore (1)   2020 Jul 31, 10:23pm     ↓ dislike (0)   quote   flag      

mell says
Communist regimes are the lucky ones with covid 19, where every deterrence is mandatory.


I'm starting to believe that deaths in China really are that low.

Because they use HCQ early in the course of the disease.
39   mell   ignore (6)   2020 Jul 31, 10:42pm     ↓ dislike (0)   quote   flag      

Patrick says
mell says
Communist regimes are the lucky ones with covid 19, where every deterrence is mandatory.


I'm starting to believe that deaths in China really are that low.

Because they use HCQ early in the course of the disease.


Could very well be. And they probably execute you if you violate the 24 hr curfew at an infectious spot.

about   best comments   contact   one year ago   suggestions