wondering when your area’s Covid-19 peak is predicted to occur?

HU-5213567

this site models the US and individual state peak estimates and projected death and hospitalization. and, if your area currently has enough beds, icu beds, and vents.


http://www.healthdata.org/covid

SaveComment275Like8
Comments (275)
Thank you for reporting this comment. Undo
maifleur03

This area is suggested to be April 20-24. Since I am on the state line I used both. Because there are still very few people with it I think the peak will probably be two weeks later and last longer than suggested.

Save    
Thank you for reporting this comment. Undo
nhb_6

Very interesting info - thank you.


Also, here is the link that I've been using for a while to track current stats. in US:

https://infection2020.com/

I think for US it is better than John Hopkins - easier to see totals and breakdown by county

4 Likes Save    
Thank you for reporting this comment. Undo
Kathy

We need to pay attention to what we need to do to keep the peak as low as possible. Otherwise———-

Save    
Thank you for reporting this comment. Undo
Stan Areted

Thanks for posting info, zeus.

2 Likes Save    
Thank you for reporting this comment. Undo
lurker111

Texas is projected to have 17- 23 deaths tomorrow. I hope not. We'll see how accurate it is tomorrow. We have 37 deaths total right now.

Save    
Thank you for reporting this comment. Undo
olliesmom

Thanks for posting!


2 Likes Save    
Thank you for reporting this comment. Undo
HamiltonGardener

Am I reading it wrong?

It says the deaths for the entire USA expected to be 2,341 by April 13

Its already past that.


eta: yes , I am reading it wrong. That’s deaths per day

Save    
Thank you for reporting this comment. Undo
elvis

When I click on the state thing, I get nowhere. Thanks anyway.

Save    
Thank you for reporting this comment. Undo
mudhouse

Here's what I did: go to home page using link above

Click on the green words in the center that say COVID-19 US State by State Projections

Click on the green bar that says United States of America

That opens a drop down menu of all the states, so you can choose yours.

****************************

Or use this link to start: https://covid19.healthdata.org/projections

Click on the green bar that says United States of America

That opens a drop down menu of all the states, so you can choose yours.

1 Like Save    
Thank you for reporting this comment. Undo
Ann

Interesting discrepancy from this chart and Cuomo's request. Looking at this chart, it says New York will need about 4000 ventilators and I'm watching Chris Wallace's Sunday show right now (from today) and one of the first reports in the show says Cuomo says he'll need 30,000 ventilators.

1 Like Save    
Thank you for reporting this comment. Undo
Kathy

Some members of Trump’s own administration were fully aware that American hospitals would soon need massive numbers of masks. On February 5, HHS Secretary Alex Azar asked for $2 billion to purchase more, and was shortchanged by the Office of Management and Budget, reports the Washington Post.

But Trump has also consistently foisted blame for his administration’s failure onto others. Appearing on Sean Hannity’s show last Thursday night, Trump denied that hospitals actually needed all the ventilators they say they need. (“A lot of equipment is being asked for that I don’t think they will need… I have a feeling that a lot of the numbers that are being said in some areas are just bigger than they are going to be. I don’t believe you need 40,000 or 30,000 ventilators. You know, you go to major hospitals, sometimes they’ll have two ventilators. And now all of a sudden they’re saying, ‘Can we order 30,000 ventilators?’”)

The dynamic with the ventilators is the same as the masks. Hospitals need way more because they have way more patients, and patients with COVID-19 are far more likely to need ventilators, and doctors and hospitals treating a highly contagious respiratory disease are far more likely to need protective gear. The fairly simple dynamic that a global pandemic creates a massively elevated demand for equipment necessary for its treatment — which is no more complex than how Halloween creates a demand for pumpkins in late October — appears to be incomprehensible to Trump.

3 Likes Save    
Thank you for reporting this comment. Undo
maifleur03

I do not think this chart is accurate especially as far as number of beds and ICU units. Missouri only shows 10.630 and the American Hospital Directory lists 16,712. According to the AHD New York has 57,261 with patients that are not in actual hospitals. These figures are from 2018 so there should be more beds available.

1 Like Save    
Thank you for reporting this comment. Undo
Kathy

La just estimated they would need 5000 ventilators. They have almost 600 on them now and it’s a lot smaller than NYC.

Save    
Thank you for reporting this comment. Undo
mudhouse

I noticed that discrepancy too, Ann, about the 30,000 ventilators Cuomo is demanding versus the 4141 that site estimates NY will need.

The number of ICU beds that site estimates New York will be short is really frightening, almost 7000 ICU beds short.

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

I do not think this chart is accurate especially as far as number of beds and ICU units. Missouri only shows 10.630 and the American Hospital Directory lists 16,712. According to the AHD New York has 57,261 with patients that are not in actual hospitals. These figures are from 2018 so there should be more beds available.

Maifleur, thanks for your comment, and for leading me to the American Hospital Directory site: https://www.ahd.com/state_statistics.html

I looked up my state's hospitals on the AHD website too, and it lists the total number of staffed beds at 3917. The OP website only shows my state as having a total number of 1752 beds. (Not ICU beds, just beds.)

But I see that the OP site says "all beds available," so maybe they're adjusting the numbers down because they know that some beds are already in use for patients with medical needs besides Covid19?

In other words, they know my state has 3917 beds total, but maybe they're estimating that only 1752 of those would be available for Covid19 patients at peak time. (?)

I have no idea how many hospital beds normally in use would be freed up by my state's current cancellation of any non-essential medical procedures.

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

Kathy: La just estimated they would need 5000 ventilators. They have almost 600 on them now and it’s a lot smaller than NYC.

Well, the website in the OP estimates that LA will need 775 ventilators.

The discrepancies in all of the statistics in the news are making me crazy. The more you read the less sense things make. I know these are estimates, but look how far apart they are.

Why would a credible site designed to estimate the needs for the Covid19 crisis estimate LA ventilators at 775, but the state itself sets their needs at 5000 ventilators?

Why does the site in the OP estimate the need for ventilators in NY at 4141, and NY state itself sets their needs at 30,000 ventilators?

1 Like Save    
Thank you for reporting this comment. Undo
Kathy

Right now La has 575 on ventilators. That was on TV tonight. Do you really think they will only need approx 200 more before tnis is over? Ho long do people stay on a ventilator on the average before they recover enough to breathe on their own or die?


eta

Part of the reason, said New York Gov. Andrew Cuomo, is that coronavirus patients are being hooked up to the breathing machines for up to 21 days, CNN reported.

Most coronavirus patients are usually on ventilators for one to two weeks, Dr. Joshua Denson, a pulmonary medicine and critical care physician, told NBC News.

Read more here: https://www.sacbee.com/news/coronavirus/article241577996.html#storylink=cpy

Save    
Thank you for reporting this comment. Undo
maifleur03

mudhouse from what I have looked at this evening comparing to what the site shows my conclusion is that it is more of a feel good site than reality site.

Save    
Thank you for reporting this comment. Undo
Lars(Z11a, Sunset 24)

God forbid that we should have too many ventilators - after all, money is more important than lives, and how would we ever sell the excess ventilators? Plus, the countries that might need them obviously don't deserve them - USA first!

6 Likes Save    
Thank you for reporting this comment. Undo
Moxie(Z4 St. Paul, MN)

Accuracy of mathematical models depends on accuracy of historical data and the underlying assumptions used to build the model. Both can be flawed. Models never capture all of the variables. Think about the models for weather forecasting and predicting hurricane behavior.

The historical data for the last few days doesn't match what my state's Dept. of Health reports. Our number of deaths is higher than what is shown. It's possible that actual data is not collected daily and the model rerun to reflect the actuals. There may be issues with the model as well. That doesn't mean that the models aren't useful. They are, especially for states that aren't in crisis mode yet.

ETA: Given an estimate for critical resources by a person in the middle of the crisis and an estimate by a brand new model, I lean heavily toward trusting the human. Actually, I'd probable use the larger estimate whether it came from the human or the model.

4 Likes Save    
Thank you for reporting this comment. Undo
foodonastump

Cuomo’s 30,000 is for a worst case scenario, not what’s predicted. I wouldn’t fault him for wanting to be prepared for the worst.

There are a lot of different calculations. My experience, I tend to believe Gov. Cuomo,” Fauci said.

4 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

Do you really think they will only need approx 200 more before tnis is over?

Please don't think I'm defending any of the numbers in the OP website, Kathy. I have no idea how they arrived at those numbers or if they're accurate. I'm just struggling to understand why these numbers are differ so much, when lives are in the balance.

2 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

Moxie: I'd probable use the larger estimate whether it came from the human or the model.

Food: Cuomo’s 30,000 is for a worst case scenario, not what’s predicted. I
wouldn’t fault him for wanting to be prepared for the worst.

I don't fault any governor for wanting to be prepared for the worst, either.

My guess is that governors over-request because they know they'll only receive a subset of that request. The problem, of course, is if every governor over-requests to be completely prepared for their worst-case scenario, and if the resulting total requested is larger than the number of ventilators that currently exist in the US supply, their desire to be ready for their worst-case scenario could cause other states to be unprepared for their best-case scenario.

It's a terrible dilemma when lives are in the balance. If the new model in the OP site is close to accurate, they estimate New York's peak resource need in 7 days, and Louisiana's in 9 days.

And as Kathy points out, reports are that Covid19 patients may need to stay on the ventilators longer than for some other illnesses, which cuts down on the effectiveness of plans to move ventilators from one state to the next one in need.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

Okay, the model is junk. Yesterday, (10 hours ago) it said Texas would have 17- 23 deaths today. Now it says we'll have 5-9 deaths. However, 5-9 is more believable than 17- 23. Our peak is estimated to be on May 6, with 28-199 deaths on that day. That's a pretty wide spread.

Also, yesterday it said that our peak would be in 17 days. Today, it says the peak will be in 32 days. It said Texas would be short 500-1000 ICU beds, now it says we'll have about 600 extra ICU beds. "All beds needed" was over 19k, now it's just under 11k.


Save    
Thank you for reporting this comment. Undo
artemis_ma

I prefer looking at the breakdown by county in my state as opposed to the breakdown by entire state. The Boston side is hitting severity first. Which doesn't mean we are not due to catch up at some point in my county, The state has 5000 confirmed cases, my county has 37 (and to date, no deaths). I am certain these numbers are severe underestimates of actual disease, considering that testing isn't universally available anywhere. The neighboring county (where most of my shopping occurs) has 151 confirmed cases, 3 deaths. Yes, we are much less populated than the eastern side of the state. I don't know per capita ratios.

Looking at the projected peaks of Massachusetts (April 14th) and the neighboring state of Vermont (April 9th) from the OP's link, and coming up with a rough average for extrapolating my area, I'd say this projects between there. Not sure how much one can trust the numbers considering that asymptomatics aren't tested to begin with.



1 Like Save    
Thank you for reporting this comment. Undo
artemis_ma

Okay , I found out that if I click on the actual map, my county has a confirmed percentage of 0.023 confirmed infection; the Berkshires (where I shop) has a rate of 0.12%. NYC has a rate of 0.4%, Boston of 0.12% just like the Berkshires.

Save    
Thank you for reporting this comment. Undo
foodonastump

Lurker - A while back I made a similar comment to Hay about the betting markets constantly changing. His response was something to the effect of it adjusting to the latest circumstances. Clearly that would apply here. What would be truly useless would be if they didn’t adjust their predictions along with updated knowledge and data.

http://www.healthdata.org/covid/updates

  • More data
    • 255 location-days of COVID-19 deaths and cases in the US have been added. What does this mean? All available data points on COVID-19 deaths and cases for 50 states plus D.C. from March 25–29 are now included in the current estimates.

Personally I don’t think the difference between 5-9 and 17-23 is very large, But then, that’s coming from NY where we’re expecting 277 today.

Save    
Thank you for reporting this comment. Undo
lurker111

Modeling isn't betting. If it's a rolling model, (by the hour) then what's the point? Just use the actual data.

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

What’s the point? Preparedness based on the latest data. Make it up as you go along based on just what happened yesterday and before isn’t particularly helpful. Ask Trump.

Save    
Thank you for reporting this comment. Undo
Ann

Mudhouse, same discrepancy with ventilators and my state. I didn't bring it up yesterday because I can't remember where I got the info, but when Roxsol and I were looking into ventilators in Alberta an Colorado on a different thread, Colorado was supposed to need about 7000 ventilators (I think it was local Colorado news I got that figure from, thus maybe from our local officials?). In any case, this website has a much smaller number.

The NY "needs" and still "needs" situation is puzzling Trump and I get that. If they had 300,000 masks delivered from just one manufacturer (who spoke yesterday) and no one knows where they went and if they have ventilators that have been delivered and are not "needed yet", when their peak is expected to be very soon, what's up? At a time like this (for the whole country and whole world), it would certainly be inappropriate for NY to be stockpiling for future medical emergencies. That should have been done previously, shouldn't be a focus at present in the midst of this crisis, and it should be determined on whose dime that should occur when the time is right (federal vs. state funding). I think it would make sense for some looking into the situation to happen, maybe just by good reporters as Trump mentioned. Maybe everything is just fine and all the federal resources being poured into the state are being put to immediate and good use (and for this particular crisis), but I'm growing curious.

2 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

lurker: Also, yesterday it said that our peak would be in 17 days. Today, it says the peak will be in 32 days. It said Texas would be short 500-1000 ICU beds, now it says we'll have about 600 extra ICU beds. "All beds needed" was over 19k, now it's just under 11k.

Thanks for posting that, lurker. I checked the peak dates for states that I was looking at last night for family members, and overnight one was moved up by almost a month, and my state's peak time was moved out by a week. My state didn't make any changes to any policies yesterday.

I assume these models change as policies and data change, but those overnight changes you documented for Texas are huge. (Throws up hands)

Last night they were predicting over 88,000 deaths total for the country. Today it has been revised down to 82,141, but the estimations update page implies changes might reflect additional states adopting social distancing measures as of yesterday.
http://www.healthdata.org/covid/updates

2 Likes Save    
Thank you for reporting this comment. Undo
lurker111

JHU said Texas had 37 deaths total yesterday, now it says 34. Why bother?

Save    
Thank you for reporting this comment. Undo
mudhouse

Foodonastump, I agree with you completely that Lurker's noted Texas change in today's deaths (17-23 down to 5-9) isn't very large. And I get it that updated data changes models.

But I can't wrap my mind around changing from a shortage of 500-1000 ICU beds to an extra 600 ICU beds overnight. What could cause that?

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

...but the estimations update page implies changes might reflect additional states adopting social distancing measures as of yesterday.

Excellent point, mudhouse, the estimates will change as the response changes. Have to adjust for that fluidity.

Save    
Thank you for reporting this comment. Undo
mudhouse

I guess it comes down to what Dr. Birx said. "Models are models."

They're not tools we can really rely on for accurate predictions, even if we'd like them to be. I guess they're better than nothing, but for a layperson like me, it's confusing.

Save    
Thank you for reporting this comment. Undo
mudhouse

Yup, food, but the ICU beds change in Texas...that's just nuts, overnight.

1 Like Save    
Thank you for reporting this comment. Undo
Tilly Teabag

Is Hay alright?

Save    
Thank you for reporting this comment. Undo
lurker111

Just don't rely on anything more than 1 hour in advance. Everything beyond that is meaningless.

3 Likes Save    
Thank you for reporting this comment. Undo
Ann

FOAS: "Lurker - A while back I made a similar comment to Hay about the betting markets constantly changing. His response was something to the effect of it adjusting to the latest circumstances. Clearly that would apply here."

Huh? The change in betting markets only has to do with the actual bets being placed. I'd certainly guess that a person's bet in a political arena is based on the latest circumstances, exactly like sports betting is.

Lurker: "Modeling isn't betting. If it's a rolling model, (by the hour) then what's the point? Just use the actual data."

Agreed.

1 Like Save    
Thank you for reporting this comment. Undo
Ann

"What’s the point? Preparedness based on the latest data."

If it's changing a lot and by the hour, that defeats the purpose of providing any useful data for preparedness.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

Never, NEVER trust anything you read from .org websites.

Save    
Thank you for reporting this comment. Undo
dadoes

The modeling map linked above says my county has 3 cases. The local OEM posted an update yesterday repeating a reference that we have 5 as of 3 days ago, and they're awaiting information on another, which would make it 6. Double what that map says.

Save    
Thank you for reporting this comment. Undo
foodonastump

overnight

As I understand it, “overnight” only because they released the latest version last night. Previous one was four days ago per the updates page. If we start seeing updates “by the hour” as lurker is alleging, with regular huge swings, then I’d take less stock in what I’m already accepting is essentially just a best guess.

Save    
Thank you for reporting this comment. Undo
foodonastump

Never, NEVER trust anything you read from .org websites.

LOL, but if you fall ill see your nearest hot tub salesman.

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

Food, I don't know when this model was released, so they may be getting more accurate data and updating for a while. For example, maybe facilities in TX could have contacted them to let them know they were using inaccurate numbers of ICU beds available. (?) I dunno.

I'll keep watching it for a while to compare the stats released daily by my own state. I'm keeping a list of those anyway, for my own sanity (or lack of sanity.)

1 Like Save    
Thank you for reporting this comment. Undo
Ann

Yeah, I just looked up my state and New York and very significant changes to both overnight. I have to lean toward the junk conclusion. Too bad, it seemed like it could be a very useful link.

Thanks Zeus, sincerely.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

You can look at the day by day projections and see that it's a very simple algorithm. I wouldn't depend on it for anything.

Save    
Thank you for reporting this comment. Undo
Kathy

The only conclusion I make from the data is they are projecting trends based on their statistics. That could change every time the numbers are updated.

Save    
Thank you for reporting this comment. Undo
foodonastump

Mudhouse - you’re right, I misread the update page as implying the prior release was March 26. That’s not what it says.

I’ll keep screenshots of NY. I didn’t remember enough of the data to note changes, aside from the peak moving out a couple days.


2 Likes Save    
Thank you for reporting this comment. Undo
Nana H

Agreed the forecasts numbers are going to change based on the daily hard data plus other changing circumstances such as travel restrictions, shelter in place orders and the like that are part of the models.

Save    
Thank you for reporting this comment. Undo
maifleur03

mudhouse and others, I think I saw a news article that the National Guard in Texas was building makeshift hospitals. If that is the case it may explain the difference. Several states are doing the same. Another thing that might have happened was if hospitals are like most in this area rooms that had been turned into singles for HIPPA reasons may now have two beds.

Edited to add that because many areas are just now seeing their first cases there was no way that the peak for most areas would be in the middle of April so a peak in May or even June for some areas should have been predicted.

Save    
Thank you for reporting this comment. Undo
catspa_zone9sunset14

I have to lean toward the junk conclusion.

This statement is a classic example of a phenomenon I point out to my students all the time, most recently with reference to "junk DNA" (another favorite being plant uptake of organic forms of nitrogen). Throughout history, many things that are not understood, usually due to complexity or lack of technology to investigate them, tend to be dismissed as being of no importance or "junk" when, it very often turns out, they are not. The data and parameters around these questions are changing hour-by-hour and there is no way that there is going to be a fixed prediction.

3 Likes Save     Thanked by HU-5213567
Thank you for reporting this comment. Undo
mudhouse

Maybe so, maifleur. Things are very fluid. I haven't been able to figure out from my reading if the beds in military-built temporary hospitals meet the qualifications to be characterized as ICU beds (?) Very good point about hospitals possibly adding beds to existing ICU areas.

Another possible problem for the model is it assesses the estimated cases/deaths based on the social distancing practices currently employed by each state. Right now this model shows non-essential businesses are still open in Texas. But while the state apparently hasn't imposed a state wide order yet, I think all of the major Texas cities have closed at least some non-essential businesses. Looks like Houston, San Antonio, Dallas, and Austin have. The model may be incapable of being adjusted for a "partial shut down" of businesses, without an official state-wide order, and that would give them a higher calculation of cases than may be truly unfolding.

If I have any of this Texas info wrong, maybe lurker can weigh in. We lived in Texas for decades, but I haven't followed their news during this crisis.

Save    
Thank you for reporting this comment. Undo
maifleur03

The few ICU units I have seen can only hold one bed it is the normal rooms that could add beds.

I am wondering if some of the HIPPA rules should be abandoned for now. The reason for going to single bed rooms here was because of the privacy rights of patients.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

I really haven't been paying attention to many counties. I'm not aware of any that were issued overnight. This model uses a very, very simple algorithm. It's easy to see and figure out if you look at the Texas deaths graph and follow the day by day count...+3 deaths per day, then +4 deaths, then +5, +6, +7...Following the curve.

It isn't worth reading beyond the current time. I had it bookmarked and already removed it.

Save    
Thank you for reporting this comment. Undo
mudhouse

Thanks Lurker.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

After looking at it again, I don't think this is a model. It looks like graph drawing software, where you enter the data and it draws a nice little graph for you. I don't know where the estimates come from. It might say somewhere on the page.

Save    
Thank you for reporting this comment. Undo
Thank you for reporting this comment. Undo
mudhouse

lurker: I don't know where the estimates come from. It might say somewhere on the page.

I stubbornly still have some hope for this site being better than nothing, because of who's behind it. (But I'm not discounting the crazy adjustments to numbers we saw overnight, as documented here.) It's just that this site, unlike some others, appears to have credible people behind it, instead of mostly political hacks. From my reading:

IMHE, Institute for Health Metrics and Evaluation, is a research center focusing on global health statistics, and part of the University of Washington's school of medicine. It started in 2007. Main supporters are Bill Gate's foundation and the state of Washington. They've had funding grants from the CDC. Two years ago they had a staff of 361 scientists and researchers (more than the World Health Organization.) They've had some conflicts with WHO because they operate independently, and WHO has seen them as competitors. The two organizations have disagreed on some past statistics.

The Covid19 model was released last Thursday. They developed the tool to help hospitals make better judgements about when their need for resources would peak.

For this Covid19 model, they're not using data about the number of positive tests. That's probably good, because states are testing at all different levels, and we don't know yet how widespread it is in the public, and that skews attempts at mortality rates. Instead, they use observed death rates for their data:

These models are based on observed death rates, and so are not influenced by differences in testing. This means that changes in death rates would alter the model, but changes in the number of observed cases, or in how states are testing, would not. We believe that in settings where testing is in relatively short supply, the sick and the very sick are getting tested; this is why we utilize deaths, as those patients are more likely to have been tested.

The data being used in these forecasts come from local and national governments, hospital networks like the University of Washington, the American Hospital Association, the World Health Organization, and a range of other sources. The model is updated regularly as new data are available, in order to provide the most up-to-date planning tool possible.

http://www.healthdata.org/covid/faqs

Save    
Thank you for reporting this comment. Undo
momofdoxies

"The number of ICU beds in the hospitals are changing because they are converting
regular rooms into additional ICU rooms. Since they cancelled elective
surgeries, many of their regular rooms are not being used." Sent to me by my DH who is the director of hospital emergency preparedness for one of our cities hospitals. That info changes daily and is sent to the state health department. They then forward it on to national agencies. Don't know if it helps to understand the differing numbers or just makes it muddier for you.

7 Likes Save     Thanked by HU-5213567
Thank you for reporting this comment. Undo
mudhouse

Thanks so much momofdoxies, that makes complete sense, and it does help explain why the number of ICU rooms would be fluid. It's encouraging that it's possible for some hospitals to convert regular rooms into ICU rooms.

Please send our thanks to your DH for his work, and stay safe.

5 Likes Save    
Thank you for reporting this comment. Undo
maifleur03

momaofdoxies ask your husband exactly what changes are being made, please. Other than hardwiring some things such as monitors and making the rooms visible from the nurses station there is very little that I can think of that would change a regular room to an ICU. Granted it has been a year and a half since I saw one.

Save    
Thank you for reporting this comment. Undo
momofdoxies

In our hospital, every patient room, with the exception of mother/baby, was designed exactly the same including the ICU rooms. Every room has observation windows and was hardwired when the hospital was built 16 years ago. There is one nurses station for every 12 or so rooms. The only change is staffing. Of course you need higher skilled nursing staff and the ratio should be 1 to 1. Not all hospitals can convert so easily, but they can convert OR suites into ICU rooms as well as surgery recovery rooms. Even the anesthesia equipment can be converted into vents quickly. And with that, DH is now on his last conference call of the night. Will try to answer any other questions tomorrow.

4 Likes Save     Thanked by HU-5213567
Thank you for reporting this comment. Undo
Tilly Teabag

My first husband died in an ICU. At least I was able to be with him as they turned off the machines. This is so sad for patients and families.

4 Likes Save    
Thank you for reporting this comment. Undo
maifleur03

Thank you momaofdoxies

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

Checked against yesterday, today’s national numbers haven’t changed significantly. Nationally there’s a slight reduction in both beds needed and shortages, the opposite for NY. Their respective peak dates remain the same.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

It's a junk website. I wouldn't use it for anything. I can find more reliable info from the usual sites. This is today, scroll up for yesterday. ^^^ Now, our peak is in 34 days.

Save    
Thank you for reporting this comment. Undo
foodonastump

Well it’s one of the models that the White House is watching, and saying it arrives at the same numbers of other models they’re following. Here’s an article about it, discusses the weakness of any modeling.

https://www.cnn.com/2020/03/30/health/coronavirus-us-ihme-model-us/index.html

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

What are “the usual sites” and what makes their predictions better than any others? Determining whose are better will be more objective in hindsight.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

They all use the same data. This one all over the place and is good for nothing other than scaring people. The others just give the actual data without making radical daily predictions. Anyway, if you like this website, please use it. Enjoy.

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

I’ll spare everyone by not belaboring the point. By now most of us are familiar with what that non-answer means.

2 Likes Save    
Thank you for reporting this comment. Undo
lurker111

It's junk science. It's already been proven beyond any reasonable doubt. Just look at the results from day to day. Not acceptable by any standard.

Save    
Thank you for reporting this comment. Undo
foodonastump

LOL

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

I suppose meteorology is junk science, too, since predictions change as input data changes.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

:^)

It will be fun, posting their failures on a daily basis.


The most definite degree of certainty
can be expressed with will and won't, followed by must and
can't/couldn't, should and shouldn't, may and may not, and might/could
and mightn't, which express the least definite degree of certainty.

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

I was an art major, and never took a course in statistics in my life (probably shows, right?) But I'm starting to understand that sites like this one, that attempt to make predictions as a helpful tool, are different animals from the other sites we use like Johns Hopkins, worldometers, our state DOH sites, etc.

The latter sites report things that have happened, and stop there. Models like this one go a step further by taking that info, combining it with mathematical formulas, and trying to extend that existing data out into the future. I don't like surprises or unpredictability, and I wish sites like this one could tell me what will happen, but they can't. They only make a mathematical guess about the future, using the data from the deaths that have already happened.

It's just a tool that hospitals can use to give them a rough idea of how well they're prepared, and it wasn't designed to give any of us comfort. With the numbers we're seeing, information for hospitals that's even slightly better than nothing can save some lives.

Some models are better than others. They can mislead people, accidentally or on purpose. A lot of the controversy over climate change is fueled by models that people disagree with. There are definitely other models that are designed to scare people to further political agendas, but I don't believe that's what this one is.

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

It will be fun, posting their failures on a daily basis.


It would be interesting to compare to your more reliable info from the usual sites.


I can find more reliable info from the usual sites.


Especially if they existed. Prove me wrong.


1 Like Save    
Thank you for reporting this comment. Undo
how2girl

Experts hopeful Australia is turning corner on coronavirus as Health Minister declares curve is flattening


https://www.abc.net.au/news/2020-03-31/coronavirus-flattening-the-curve-greg-hunt/12107404




3 Likes Save    
Thank you for reporting this comment. Undo
Nana H

Let's hope so for Australia! It's been a rough more than few months between this and fires.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

Why should I waste the time to prove you wrong when you know I'm right? It's a simple algorithm that I could do in my head.

Example: 20k cases = 2k beds = 300 ICU beds.

BOOM goes the dynamite! :^)

Save    
Thank you for reporting this comment. Undo
catspa_zone9sunset14

The Kinsa study that is tracking thermometer readings across the U.S. is showing that social distancing policies have significantly decreased infection rates (both flu and Covid) where they are in place -- very interesting source.

Too bad the state government in Texas has yet to mandate either shelter-in-place or closure of non-essential businesses state-wide. When exponential dynamics are the case, even small decreases or increases significantly change the curve, so Texas may essentially be choosing to increase the toll and lengthen the epidemic, inflicting unnecessary death and grief on its citizens.

2 Likes Save    
Thank you for reporting this comment. Undo
foodonastump

Why should I waste the time to prove you wrong when you know I'm right? I


What I “know” is that you repeatedly waste a lot more time posting about why you won’t provide evidence, than it would take to simply post a link or two to your supposedly better, readily available information. Tells me all I need to know. Show me a better model that we can track simultaneously, or admit by implication that they’re imaginary.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

How about I just explain to you that I never mentioned a "better model". I mentioned better sites that don't offer the radical predictions.

Here's the quote that you must have misread.

"The others just give the actual data without making radical daily predictions."

:^)

Save    
Thank you for reporting this comment. Undo
foodonastump

Oh so your brilliant observation is that reporting the past is more accurate than predicting the future. Thanks, good to know.

3 Likes Save    
Thank you for reporting this comment. Undo
lurker111

I know, you wished you could have figured it out on your own. :^)

Save    
Thank you for reporting this comment. Undo
dadoes

Any data that doesn't follow with one's personal position is junk data ... as Trump continually makes clear.

The map is updated for my county, which is now 7 cases.

2 Likes Save    
Thank you for reporting this comment. Undo
catspa_zone9sunset14

I guess you're more inclined to be a historian than a scientist, eh, lurker?

1 Like Save    
Thank you for reporting this comment. Undo
Kathy

I found this chart interesting and posted it on another site. Masks + social distancing improve the curve.

Save    
Thank you for reporting this comment. Undo
mudhouse

https://archive.is/zQbDq

Here's a link to the NY Times article linked in catspa's comment above, for folks who don't subscribe or would rather not sign up for emails.

I have so many thermometers I haven't been in the market for years, and I'd never heard of Kinsa thermometers until Annie mentioned their map recently. Interesting how that technology captured some of the earliest national evidence of increased fever caused by the virus.

It's encouraging to see that the fever data can be overlayed with the dates of social restrictions and business closings to show that those actions make a difference.

Save    
Thank you for reporting this comment. Undo
lurker111

Let's use Cuomo's predictions..."10 million cases" = 1 million beds = 150k ICU beds = 120k ventilators. He needs to get busy!

I guess you're more inclined to be a historian than a scientist, eh, lurker?

I don't talk science with crats. History, either.

Save    
Thank you for reporting this comment. Undo
HU-5213567

this is the model presented in todays briefing.

Save    
Thank you for reporting this comment. Undo
Ann

I heard about it today. The link in the OP gives me an error now. Does it still work for others?

1 Like Save    
Thank you for reporting this comment. Undo
HU-5213567

new link. https://covid19.healthdata.org/projections. maybe a poster higher up can put the new link in. i cannot edit my post.

Save    
Thank you for reporting this comment. Undo
foodonastump

Original still works for me.

Save    
Thank you for reporting this comment. Undo
Ann

Thanks, I just found it.

I was feeling like it was junk (up thread), but it does appear it is being used heavily and is highly respected. Sorry for my too soon judgement.

2 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

The original works for me too, but I edited my post up thread to include your latest link too Zeus, since it eliminates one click.

2 Likes Save    
Thank you for reporting this comment. Undo
Ann

I hope my state lives up to its projections. If so, we're in for a rough couple weeks right now, but down to practically no additional deaths 1-2 weeks into May.

I heard Virginia has been shut down for many weeks and that makes sense based on their graph, peaking much later than mid-April.

This makes me think it could eventually make good sense to open some areas of the country much sooner than others.

2 Likes Save    
Thank you for reporting this comment. Undo
Kathy

Michigan should hit it’s peak in 15 days, somewhat earlier than some other states. How long after the peak do you think it would be wise to start opening up?

Save    
Thank you for reporting this comment. Undo
HU-5213567

lurker111

It's a junk website. I wouldn't use it for anything. I can find more reliable info from the usual sites. This is today, scroll up for yesterday. ^^^ Now, our peak is in 34 days.

=====

LOL - maybe you should call Trump.

1 Like Save    
Thank you for reporting this comment. Undo
Nana H

There is a difference between the peak and an all clear. Reasonably the length of dangerous times after the peak is as long, or longer , than the time proceeding the peak . In other words the days after the worst will be the same as the days before it . I worry people will let their guard down when the numbers start to go down.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

I think it's funny how the crats accept a model that makes Cuomo look like a clown, and one that Trump uses. :^) It's a simple math equation. If it works, cool. I think it's too soon to tell. These type of models will never be wrong, because they adjust the numbers daily to fit the model. If it were a real model, it would have to have some degree of certainty.

Save    
Thank you for reporting this comment. Undo
Kathy

Yes, I did hear Birx say the down slope will be longer than the up. So I guess we can estimate from there.

lurker, it’s all projection. The truth is in the pudding and lots of variables, imho.

Save    
Thank you for reporting this comment. Undo
Ann

"How long after the peak do you think it would be wise to start opening up?"

If I were an expert on pandemics, I'd be happy to give you an answer:)

1 Like Save    
Thank you for reporting this comment. Undo
Kathy

I was only asking for an opinion. If we were experts we wouldn’t listen to Trump.:)

1 Like Save    
Thank you for reporting this comment. Undo
Ziemia(6a)

About the lower death rate - over the weekend -

https://www.washingtonpost.com/graphics/2020/national/coronavirus-us-cases-deaths/

********

In several places - when the curve flattened just a bit - people became less vigilant and the lower rate of new cases reversed (↑).

2 Likes Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

The linked data for California doesn't reflect what California's own projections are for coronavirus in the state.

First, Governor Newsom says that the state's projections for peak cases is during May -- long period for differences between start of cases in Northern and Southern California. The site predicts April 27.

Second, number of ventilators projected to be required -- 10,000 -- are much higher than the linked site shows -- 1,800.

Third, the state is preparing more hospital beds in advance of the surge even as the site says "don't worry, be happy."

I don't trust the projections presented on that site. I hope that they are correct, but I doubt it.

Save    
Thank you for reporting this comment. Undo
Nana H

Misplaced I suppose but just heard a good one on Vision TV.......

When will we have another chance to save humankind by sitting on our coach watching TV.. ...STAY HOME

2 Likes Save    
Thank you for reporting this comment. Undo
jerzeegirl (FL zone 9B)(9b)

Thanks for posting this Zeus. That is a great website.

Save    
Thank you for reporting this comment. Undo
Kathy

Ziemia, I believe vigilance is key as you stated. That will be very hard to maintain but neccessary.

Save    
Thank you for reporting this comment. Undo
HU-5213567

nancy - the model is the projected curve and is based on available data. it will change every day as the hard data is known. state leaders need to be prepared to be a top outlier, which it sounds like your state is.

2 Likes Save    
Thank you for reporting this comment. Undo
HU-5213567

so sad to see the model shows added 12k projected deaths since sunday. please do your part and maybe we can bring the number down.

today 93, 765

it was 81,114 on sunday


1 Like Save    
Thank you for reporting this comment. Undo
Annie Deighnaugh

Latest tally in our county: 1,870 cases, 38 deaths, and no one has recovered. One doctor said, the only way we are clearing ICU beds is if someone dies.

Save    
Thank you for reporting this comment. Undo
lurker111

Our county will probably skyrocket this week due to the in home testing starting today. Texas will probably catch NY in no time. Our county doubled our cases yesterday. It will be really bad by next week. My city had 1 active case a couple of days ago, we had 9 yesterday, and today will probably be much worse. We've had 10 recoveries and no deaths. 3 have recovered in our city.

If anyone is interested...

https://hayscountytx.com/covid-19-information-for-hays-county-residents/?fbclid=IwAR2l-KjidCV_ZcF-uBZNf6rLpYgLoew1tXGX7Mk6LTqNm9flg7a4oukCFaY

1 Like Save    
Thank you for reporting this comment. Undo
gyr_falcon(Sunset 23)

A church just sent me this flier. I assume they didn't intend their graphic to resemble the site graphs, with crosses for the burials at the peak of the pandemic... I'm hoping for a better Friday than that.



1 Like Save    
Thank you for reporting this comment. Undo
Ann

The website is supposed to be updated today, but it hasn't been yet. Last update was on 4/1.

1 Like Save    
Thank you for reporting this comment. Undo
KennsWoods formerly kentrees

Last I heard FL will peak early-mid May. It's gonna be a long spring in the land of the mouse.

Save    
Thank you for reporting this comment. Undo
mudhouse

Yup Ann, apparently no update on the site today, even though it still says to check back today for the next update.

For what it's worth, their predicted death rate for my state, NM, is running close to the actual numbers since I started checking daily stats.

But their predicted death rate for another state where I have family, Alabama, is running high. The actual death rate so far is running less than 25% of their predicted cumulative total. I'm hoping that means they might revise the total predicted deaths down for that state, but I know it's still early.

1 Like Save    
Thank you for reporting this comment. Undo
maifleur03

According to the first chart my area is supposed to peak in May carrying over into July. There will be many peaks and valleys until this plays itself out.

Save    
Thank you for reporting this comment. Undo
Ann

Mudhouse, the actual death rate in my state has been half the daily predicted death rate for the last couple days. Yet, Birx talks about my state as one they are watching closely (in a concerned way rather than a pleased way), so I'm waiting to see what the updated chart says. It's kind of weird the website says to check today for updated charts and yet no updated charts today.

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

For what it's worth, their predicted death rate for my state, NM, is running close to the actual numbers since I started checking daily stats.


NY, too. NY actual yesterday:

NY predicted for yesterday, as of 4/1 update:


I don’t have any specific dates highlighted from when I first started tracking on Monday, but you can see the graph hasn’t changed much:


Save    
Thank you for reporting this comment. Undo
Nana H

My God FOAS , that is terrifying!

Save    
Thank you for reporting this comment. Undo
Annie Deighnaugh

I don't know why but it seems every day for the past 2 weeks, they've said our area is going to peak in 2 weeks and they're still saying it. So I guess it'll peak in 2 weeks...we just don't know when the clock starts on that.

1 Like Save    
Thank you for reporting this comment. Undo
jerzeegirl (FL zone 9B)(9b)

Our peak in Florida is May 6-8. That seems like forever.

Save    
Thank you for reporting this comment. Undo
Ann

The charts didn't get updated yesterday (when the text said to check back) but it no longer has a check back date on it - so who knows what they plan anymore as far as updates.

Save    
Thank you for reporting this comment. Undo
foodonastump

Well I guess at least no one can complain that it’s changing by the hour.

Yes, Nana, it’s pretty scary. I think I can avoid going out for a few weeks. Maybe one curbside pickup for my parents. I’m glad our peak is earlier rather than later, although I don’t have a good grasp on what the way down looks like.


1 Like Save    
Thank you for reporting this comment. Undo
maifleur03

I noticed several states have changed to a peak in May rather than in the next two weeks so staying at home has increased that distance. If rather than having many states peaking at the same time and having lower peaks it could allow states that have already peaked to send equipment to other areas.

I think the number of projected deaths at the height has been lowered but overall deaths increased.

2 Likes Save    
Thank you for reporting this comment. Undo
chisue

My Sunday Chicago Tribune reports on two disparate models for Illinois. (Headline is: When will Illinois outbreak peak?) There are graphs.

IMO the story is couched to avoid panic. Only the final paragraph says deaths will be 18,000 best case, up to 64,000 worst case.

One model has a peak April 20. One says June 8 - 15.

Best case for each model depends on the public following Gov. Pritzker's declaration to stay home AND the addition of beds/medical services.

Current bed count (throughout the state, and just 'beds', not ICU beds) is listed at about 15,000. The April-peak model shows a tipping point of virus-patient need v. beds arriving within few days if the public fails to stay home. The June-peak model sets that non-compliant point at May 11, *and* it presumes we will have twice the number of beds by then.

Save    
Thank you for reporting this comment. Undo
Ann

New numbers did get released this afternoon or evening. Things are looking much better for the U.S. As an example, the expected number of deaths in my state of CO is now 302, 14% of the 2154 expected in the 4/1 version of the chart.

Save    
Thank you for reporting this comment. Undo
mudhouse

The Washington University site in the OP has updated, and for many of the states I've been monitoring, they've made some very big changes. Some predicted peak dates have changed, as well as the total number of predicted deaths. So folks might want to revisit and check the updated projections for your state.

********************

The US peak resource use date is still shown as April 15, but the total death projection went down from 93531 to 81766.

Alabama's projected total deaths (thru Aug 4) has changed radically, from 5516, down to 923. This is much more in keeping with the actual cumulative death count I've been tracking from the state website.

Texas's projected total deaths was reduced from 6392 to 2025, and the peak date has been moved up from May 6 to April 19.

Louisiana's projected peak date is now showing that Louisiana has already peaked, as of April 1.

New York's projected peak date has been moved up to peak in only three more days, on April 8.

Virginia's projected peak date has been moved way up, to April 20.

My state (NM's) peak moved up a few days, and the projected death total increased slighty to 603. That also fits with the actual number of accumulated deaths I've been getting from our state website. We'd gotten a bit ahead of what they were projecting, so that change seems to fit.

**************

I should note that in all of the above cases, I've been using the date they show for peak resource use (the first date shown in bold in each state's projection.) If you scroll down a bit, the peak death rate usually lags several days after peak resource use.

(Edit, they changed the total US deaths downward as I was typing the above, so I edited my post.)

2 Likes Save    
Thank you for reporting this comment. Undo
foodonastump

Glad your state is looking so much better. I didn’t keep the total death charts. I do see that the US daily death chart gas spiked up 20%, NY’s is up just slightly, so it’s not everywhere by a long shot. Wonder what made CO improve so drastically.

ETA obv mudhouse is tracking this more closely!

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

Cross posted with you, Ann! I'm glad things are looking so much better for Colorado, at least according to the Washington University model. That's a big change.

2 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

foodonastump, they have a page where they've explained past updates, but there's nothing new there yet. Maybe they'll add some notes tomorrow.

On the Alabama projection, they made massive changes to the resource numbers. Until today they were saying Alabama needed 27,498 beds, and only had 5743. Now they say Alabama only needs 2052 beds, so the 5743 they show is more than enough.

They were saying Alabama needed 4382 ICU beds, now revised down to needing 400.

Ventilators went down from a projected need of 3504 to 340. Big changes that I'm glad to see.

1 Like Save    
Thank you for reporting this comment. Undo
HU-5213567

excellent news. sad at the number even with the downturn, but there is still hope it can go lower. please know the model assumes isolation recommendations are being followed...

3 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

foodonastump: ETA obv mudhouse is tracking this more closely!

I admit to nervous note-taking, foodonastump, on some states where I have family. The Alabama resource numbers seemed really off to me, so I was watching those especially.

I'm sorry to hear that NY's projected total went up slightly. I had only been watching the peak date for your state. I sure have been thinking about you lately. Saying "stay safe" sounds a little trite, but it's sincere, and I'm so wishing the best for you and your family.

3 Likes Save    
Thank you for reporting this comment. Undo
maifleur03

Since some of the states that were to have peaks within the next few weeks have barely started having any deaths I find it hard to believe that the peaks will happen that quickly.

I just looked and for the states I was watching the death rate and peak times have been drastically reduced.

Save    
Thank you for reporting this comment. Undo
foodonastump

I appreciate that mudhouse, you the same. Swore I was done shopping for a while but then realized I was out of vacuum bags. Common Hoover size sold out on amazon, so I went to a hardware store. My wife read me stats for my town this evening, hundreds of confirmed cases and even my little hamlet has dozens, so it’s time to hide the car keys.

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

Yup; I guess it's our job, food. DH says we're out of oranges, so I read him the stats. He reaches the conclusion that we can live without oranges. Wise man. I'd definitely be in Hunker Down Mode in your area.

3 Likes Save    
Thank you for reporting this comment. Undo
foodonastump

I ended up throwing out my back shortly after that trip. I’m not going anywhere for a bit, will be happy just to get to the toilet in time!

1 Like Save    
Thank you for reporting this comment. Undo
barncatz

FOAS, sorry to hear that- that's so miserable.

mudhouse, I am amazed at your command of those charts. They make my eyes twirl. Luckily, that's not a symptom of the virus, but it's still uncomfortable.

I started making the doggo's food last Fall and just realized that DH didn't shop for those ingredients. We only have six more days. Ruh ro. I could off the kitty cat, I suppose, for their protein source. JK!

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

Barncatz, I saw a cartoon recently. Long board room table, full of dogs. Head dog says: "Our pandemic plan is working; we're now reaching peak time at home with our humans."

Lone cat at end of table says: "Why wasn't my department told about this?"

2 Likes Save    
Thank you for reporting this comment. Undo
foodonastump

The updates page has been updated. I didn’t try to follow the whole discussion, bottom line more data to work with and tweaks to the methodology. They happened to specifically mention the sharp change in CO vs the stability of the NY model:

Colorado

In Colorado, fairly substantial changes have occurred in our estimates for the trajectory of COVID-19 daily deaths over these three iterations of our model. Even though projected estimates all lie within the uncertainty intervals of each other, the predicted peak day for daily deaths has shifted considerably over model iterations, as have predicted peak daily deaths. As discussed above under the COVID-19 death model updates, these changes are driven by a combination of increased availability of reported death data and improvements to our modeling approaches since our first release on March 26.

New York

Unlike Colorado, New York’s predictions have been far less variable over our model iterations. This is due to New York’s reporting of COVID-19 deaths for a longer period of time, at least relative to other states, as well as having a much higher death toll. Our projections for New York have been fairly stable over our past releases and track the reported death numbers closely, highlighting our model’s good performance in settings with large epidemics and frequent data reporting.

http://www.healthdata.org/covid/updates

Save    
Thank you for reporting this comment. Undo
jerzeegirl (FL zone 9B)(9b)

It looks like Florida's peak got moved to April 21. But you are right, ann, the chart hasn't updated yet.

Save    
Thank you for reporting this comment. Undo
Ziemia(6a)

Thought I'd capture the current projection for my state


1 Like Save    
Thank you for reporting this comment. Undo
Ann

"It looks like Florida's peak got moved to April 21. But you are right, ann, the chart hasn't updated yet. "

JG, what do you mean? The charts did update and the updates posted yesterday. Mudhouse and I both noticed and posted about the updated charts last night.

Save    
Thank you for reporting this comment. Undo
Ann

The stock market seems quite pleased about the improved projections. I've spent an hour watching CNBC and they are associating the 1000 point Dow increase to the improved virus projections. How great it will be to get back to an operating economy!

2 Likes Save    
Thank you for reporting this comment. Undo
maifleur03

The charts became even stranger as the predicted number of deaths and need for beds was lowered. At 562 beds that is less than is currently being used.

Save    
Thank you for reporting this comment. Undo
Thank you for reporting this comment. Undo
Nana H

Now that is great news ! Won't help with this go 'round but it sure as hell is needed ASAP 'cause this beast is predicted to return. Great work !

Save    
Thank you for reporting this comment. Undo
olliesmom

Thanks Ann, for posting! Great to hear some good news! We need it fast-even though I know it takes some time.

2 Likes Save    
Thank you for reporting this comment. Undo
elvis

wondering when your area’s Covid-19 peak is predicted to occur?

Nope.

Save    
Thank you for reporting this comment. Undo
Ann

Wow, even better day for the stock market than earlier. I got busy and had no clue it ended up yet another 600 points on top of the 1000 point morning gain! Just looked and that's good news.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

Then

And now

It's a bad algorithm and nothing I would use for future planning.

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

yawwwwn

Save    
Thank you for reporting this comment. Undo
lurker111

:^) I know, I know.

1 Like Save    
Thank you for reporting this comment. Undo
maifleur03

I have to agree lurker because little has changed to reduce the number of beds that would be needed. Some states already have more than the suggested amount in use.

1 Like Save    
Thank you for reporting this comment. Undo
Thank you for reporting this comment. Undo
foodonastump

Maifleur - which states? Which were you referring to with 562 above?

Save    
Thank you for reporting this comment. Undo
lurker111

Some people look at a chart and see "science". Others see easy algorithms and bad methodology.

That tweet isn't fake. We have biden's video, also. :^)

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

Some people look at conspiracy sites and think they’re more than entertainment. Others see nonsense.


:^)


Save    
Thank you for reporting this comment. Undo
mudhouse

The "beds needed" numbers still bother me because the changes in many states were so huge, including ones I've really been monitoring.

But the updates page that foodonastump linked to above does provide an explanation for changes to that number. http://www.healthdata.org/covid/updates

One thing to consider is the first March 26 release used the peak death rate from Wuhan China to try to model the predicted death rates for US states. That was the only city that had peaked, at that point. They used Wuhan data about the number of deaths, and the time from the first social distancing practices to the peak number of deaths.

I don't know about you guys, but I have no faith at all in data from Wuhan, so when I saw that...hmm. But, it was all they had to use at the time in terms of a city that had reached peak death level.

Since then, seven more cities in Italy and Spain have peaked. That information was used to reshape the projections for the severity and timing of peaks in each state in the US. The more data they have, the more accurate their modeling can be.

Still, they're using European data to try to predict US death rates, and I've read that Spain and Italy may have differing methods to count which deaths were Covid19 deaths, and which were not. So, there's that problem, too, although the updates page doesn't mention it. I'm just adding my own gripey two cents on that point. I think it can be problematic to pull data from one country and apply it to another, even if the other country isn't China.

But again, they have to use the data they have, and that's all they have.

The updates page also explains that they have more data now about how long people need to stay in the hospital beds. They know how long some critically ill people live before dying in the hospital, and how long critically ill people who recover stay in ICU beds, and how long moderately ill people who recover stay in regular hospital beds. Those things, of course, drastically affect how many beds are needed. Basically, they adjusted the number of beds needed (drastically) because they learned that the time they were estimating for average stays was too long.

I'm sure there are inaccuracies; I'm not confident they can be accurate about the number of available beds that really exist in each state, especially since hospitals are apparently doing an incredible job of making changes to increase their capacity.

My own state gov talked about that in one of her pressers, after momofdoxies was kind enough to post some info from her husband on this point, earlier in this thread. She said my state has always had a terribly low per capita number of hospital beds, so NM hospitals are commonly designed in ways that allow them to expand capacity if needed. I never knew that.

Also, the model uses social distancing programs (closing schools, closing businesses, stay at home orders, etc) as a part of their calculations. But my state shows we have no stay at home order, and that's completely wrong; it's been in effect since roughly mid March, but our state law doesn't use the "stay at home" wording. I'll bet other states have similar complaints with the model accuracy on that point.

Lots of moving parts. With all due respect to my good friend lurker, I think it's a lot more complicated than saying "it's a bad algorithm." I'm still in the camp of those who think the model (for me, anyway) is more helpful than nothing.

Models aren't perfect. :-) They're just models.

3 Likes Save    
Thank you for reporting this comment. Undo
lurker111

It's junk science. Much like the global warming models that all fell flat. This model falsifies itself every time they adjust the numbers.

Save    
Thank you for reporting this comment. Undo
lurker111

Wow, even better day for the stock market than earlier.

That's why I wasn't here much yesterday. :^)

1 Like Save    
Thank you for reporting this comment. Undo
maifleur03

One of the oddities I noticed is that my state Missouri only needs 562 beds population 6,137,428 while Iowa 726 population 3,155,070. I would think that a state with twice the number of people would expect to have the need for more beds.

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

Good question. Not just higher population but higher population density, and significantly lower health index. (Based on quick search.)

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

This model falsifies itself every time they adjust the numbers.

Models are just tools, not crystal balls. With a model like this one, trying to assist hospitals in a health crisis, the goal isn't to be true/false or right/wrong; it's to shine a flashlight a bit further down a dark hallway to help with planning.

Every time they adjust the numbers, using new data about known things that have actually happened (like deaths) the model's accuracy increases.

If we reject the use of models because they can't flawlessly predict the future without refinement as new data presents, we're eliminating a tool that can save lives.

I get it that adjusting the numbers along the way can seem like cheating, if we think of models as a one-time guess about the future, and we think the guess is either right or wrong.

But I think of it like adjusting the flashlight beam every thirty feet down the hallway. While it's true we'd get to the end of the hallway without it, the idea is to try to get there with fewer mistakes. And for medical resource shortages, mistakes mean losing lives, and horrible tolls on the medical professionals who fight to save lives.

4 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

I do completely agree with you, lurker, that models can be wrongly used to achieve politically popular agendas, like any tool. I saw one model about Covid-19 that I thought was shameful. I just think this particular model is a valid attempt to be helpful, and I don't think we should throw out the idea of all scientific models just because some are manipulated and misused.

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

That is a good question, maifleur. I wish we understood more about how the researchers arrived at their estimation of beds needed; that's bothered me from the start, as a layperson trying to understand this.

1 Like Save    
Thank you for reporting this comment. Undo
catspa_zone9sunset14

One of the oddities I noticed is that my state Missouri only needs 562 beds population 6,137,428 while Iowa 726 population 3,155,070.

Read the top bar showing measures implemented, maifleur, which are part of the assumptions. Missouri has a stay-at-home order in place now and closed educational facilities on March 23. Iowa has no stay-at-home order and did not close educational facilities until April 4. Small differences like that can cause large effects in exponential situations.

It's junk science. Much like the global warming models that all fell flat. This model falsifies itself every time they adjust the numbers.

It's okay if you don't understand high-level numerical analysis, lurker. Most people don't, outside experimental high-energy particle physicists and such. And even DH, who is one of those, in the past was skeptical of Baysian approaches to modeling, too, though he has grudgingly come around to seeing their value when it comes to life science in particular and other complex systems.

2 Likes Save    
Thank you for reporting this comment. Undo
lurker111

A model has to have at least some degree of accuracy. You have to be able to run it in reverse and come close to the actual numbers. This isn't a model, it's just people drawing graphs based on the current numbers, and using a simple algorithm for a crystal ball. Totally useless.

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

Read the top bar showing measures implemented, maifleur, which are part of the assumptions. Missouri has a stay-at-home order in place now and closed educational facilities on March 23. Iowa has no stay-at-home order and did not close educational facilities until April 4. Small differences like that can cause large effects in exponential situations.

Thanks catspa. I understood that each state's curve size and length was affected by the distancing measures, but for some reason I didn't consider that the "beds needed" estimates were also created after looking at those measures.

I never thought I'd be wishing in my 60's that I'd taken a class in statistics somewhere in life. I used to run from trying to read a graph of any kind.

Save    
Thank you for reporting this comment. Undo
maifleur03

catspa the state of Missouri only started a stay at home yesterday as in this week. Some areas of the state have had stay at home orders longer than that but most of the rural areas have not. Since this is only the second day of the state wide stay at home it should not have made any difference.

Save    
Thank you for reporting this comment. Undo
catspa_zone9sunset14

maifleur, these are predictions based on IHME's assumptions about how these factors will affect resource needs going forward. In this case, the assumption would be that having a stay-at-home order in place will reduce future cases and resource needs than having no order in place at all, not that it already has. The four factors IHME lists at the top of each state's page are included in their model calculations because they are assuming they will affect outcomes. Here's an excerpt from their latest comments about updates for the model:

Model update 2: Weighting of social distancing policies in the COVID-19 death model.
 Our previous models had a composite measure for estimating the effects of social distancing policies on the course of the epidemic. Until today’s release, all types of mandate had been combined together by giving them equal weight. Of the four social distancing policies we currently track (i.e., school closures, stay-at-home orders, non-essential business closures, and travel limitations), no state had enacted a travel limitation mandate that met our criteria until quite recently (i.e., Alaska, implemented March 28 and incorporated into our April 2 release).Subsequently, we continue to primarily focus on the three measures – school closures, stay-at-home orders, and non-essential business closures – more widely implemented by US states to date. If or as more states mandate severe travel limitations, we will revisit its inclusion into our updated weighting approach.

And it continues, read further comments here: http://www.healthdata.org/sites/default/files/files/Projects/COVID/Estimation_update_040520.pdf

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

The IHME site added projections for 29 additional countries today; this one is for the UK.

https://covid19.healthdata.org/united-kingdom

And a link to today's explanations about the projections they've added for countries in the European Economic Area.

http://www.healthdata.org/covid/updates

1 Like Save    
Thank you for reporting this comment. Undo
maifleur03

catspa that is fine and dandy but how does putting a stay in place yesterday lower the number people who will develop the virus? There should be no difference in numbers from Sunday to Monday.

Save    
Thank you for reporting this comment. Undo
HU-5213567

UK estimate. I hope I am reading this wrong or that this is way off. UK projected 66k deaths. UK is 25% the size of the US, which is currently projected 81k.

Save    
Thank you for reporting this comment. Undo
mudhouse

I hope the projection is wrong for the UK too, Zeus, but according to this current chart on Worldometers, the per capita Covid19 death rate in the UK is now almost 2.5 times that of the US.

UK: 91 deaths per million

US: 38 deaths per million
https://www.worldometers.info/coronavirus/#countries

1 Like Save     Thanked by HU-5213567
Thank you for reporting this comment. Undo
catspa_zone9sunset14

Maifleur, they aren't updating the model daily. They did update it at some point mid-day (my time) on Sunday, April 5, if I remember right. Did you look at it Sunday morning (while it was still showing the April 1 prediction)? In that case, looking at it on Monday, after the April 5 updates were published, you might see a change, because underlying assumptions had changed. From what I read, Missouri's governor only issued the statewide order on April 3, so the April 1 predictions wouldn't have included that as an assumption, while the April 5 (and today's April 7 version) would.

Save    
Thank you for reporting this comment. Undo
lurker111

They'll get it right when it's over.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

That worldometer destroys the propaganda that the USA was far behind on testing. No one had enough tests. I'm glad that hoax has been put to rest.

Save    
Thank you for reporting this comment. Undo
jerzeegirl (FL zone 9B)(9b)

Here is a interactive map with the US death count. The first 5,000 deaths came in just over a month’s time, and in fewer than five days, the second 5,000 followed. Map

Save    
Thank you for reporting this comment. Undo
lurker111

Fake news paywall.

Save    
Thank you for reporting this comment. Undo
ubro(2a)

That worldometer destroys the propaganda that the USA was far behind on testing. No one had enough tests. I'm glad that hoax has been put to rest.

Nope:
The worldometer posts how many tests each country did it does not say when. If you started late to the game and then played catch up those figures make it look like the virus was taken seriously but the spread and the "hot spots" were well past the time for containment.

You can use the numbers to make yourself feel good, but that is not what is playing out in real time.

ETA I have been following the worldometer stats daily, and the US was well behind other countries in testing for many weeks, even the provinces of Alberta and BC had the US beat for a period of time.


2 Likes Save    
Thank you for reporting this comment. Undo
lurker111

Get real. There hasn't been enough tests globally to test half of Texas. Starting early?

Save    
Thank you for reporting this comment. Undo
mudhouse

The IHME model made another downward revision today for US deaths. They revised the total projected US deaths through Aug 4 down from 81,766 to 60,415.

New Mexico's total projected deaths went down by about 1/3.

Alabama's total projected deaths went down by about 1/3.

But Texas's projected deaths went up by 17 deaths.

(Lurker, can you please find those 17 people and tell them to stay home? Don't make me drive over there.)

3 Likes Save    
Thank you for reporting this comment. Undo
lurker111

:^) Probably illegal aliens and other foreigners.They walk around like they own the place. You can't even drive through town anymore. They all walk right in front of your car and don't even look. They just hold their heads high and act like they're better than everyone else, and if you hit them, it's all your fault. They even race to get in front of you at the traffic lights so you have to wait for them to crawl across the street. Third world mentality.

On another note, we had a scare last night. My cousin's grandchild had to go to the hospital with a fever and shortness of breath. It was a urinary tract infection. I don't know where the shortness of breath came from, probably a panic attack. Her stepfather's uncle died last week from the coronavirus, in Fort Worth.

Save    
Thank you for reporting this comment. Undo
mudhouse

I was thinking it was probably 17 of my old friends in Texas. Trying to tell some of them what to do was like trying to push a string. Big-hearted and strong-willed people. I love Texas.

So glad your cousin's grandchild is "only" dealing with a UTI. Hope he or she recovers OK. I'm sure I'd have a panic attack too, in that situation.

2 Likes Save    
Thank you for reporting this comment. Undo
Ann

Lurker, that would be a family scare.

You know, I'm almost surprised more HT posters aren't experiencing family coronavirus cases. Just a couple have written about such situations. It's kind of given me an optimistic feeling that so many people from so many states (and various countries as well) have escaped this so far.

3 Likes Save    
Thank you for reporting this comment. Undo
Ann

Great news about the drop in projected U.S. deaths. Slight increase in my state, but you might remember we had a huge drop between 4/1 and 4/5. Now, we have a slight increase. My husband heard on the news that 100 vents are headed to my state, which should just about fill our need. Not long ago, our Dem governor requested 10,000 vents, which was many times the need that was ever projected for our state. Even before our enormous 4/5 drop in numbers, our need was projected at 1052 vents (not anywhere near 10,000).

3 Likes Save    
Thank you for reporting this comment. Undo
catspa_zone9sunset14

Probably illegal aliens and other foreigners.They walk around like they own the place. You can't even drive through town anymore. They all walk right in front of your car and don't even look. They just hold their heads high and act like they're better than everyone else, and if you hit them, it's all your fault. They even race to get in front of you at the traffic lights so you have to wait for them to crawl across the street. Third world mentality.

How did you know they were illegal aliens, lurker? Did you ask or did they just "look" that way? Or maybe you know them?

Save    
Thank you for reporting this comment. Undo
mudhouse

I'm glad they adjusted the NM total down, not just for the obvious reason, but also because our state went up in the last adjustment, and the new projected running death total was already showing to be much higher than the daily stats coming in. It looked like they'd over corrected. I'm glad they're making tweaks (?) in between the major updates.

Same with Alabama. Even with the recent change downward, the projection was still running almost three times as high as the actual death rate, so it's good they tweaked that state again too.

I'd have less confidence in the model if it looked like they were ignoring reality, instead of adapting the projections.

Save    
Thank you for reporting this comment. Undo
lurker111

Looks like our peak date has been moved, also. It was May 6, then April 17, now it's April 22. They'll get it right, after it happens.

Save    
Thank you for reporting this comment. Undo
Ann

"They'll get it right, after it happens."

Yeah:) I'm watching it with interest and I don't think it's junk as I once did, but it has certainly been less than stable! At this point, I'm somewhere in the middle of agreeing with your opinion and thinking it provides some value.

1 Like Save    
Thank you for reporting this comment. Undo
HU-5213567

US projection is down to 60k. Sadly, maybe because UK began mitigation relatively late(?) UK is still 66k

Save    
Thank you for reporting this comment. Undo
llucy

I'm in OH. According to the site, bed peak is today. Death peak is the 12th. Previous peaks were the 21 st, then the 19th.

Another site I check tracks confirmed cases/hospitalizations/deaths by county. The stats in my, and surrounding counties are lower and growing slower than I would have expected. Maybe this is the result of people staying home, but I wonder what will happen as the weather becomes more enjoyable.

I live on a lake. Retired snowbirds will be returning from FL in the next month and local stats may change to reflect that. Some folks who have homes here live elsewhere and use their lake house mainly in the summer for weekend getaways and family vacations. The community population doubles, triples,...because of people having house guests and entertaining.

I want feel optimistic about summer, but I know 'the curve' here could change and I need to be prudent accordingly.



Save    
Thank you for reporting this comment. Undo
mudhouse

llucy, I think your post is a good example of why different regions will have different experiences. So many things are changeable, too. Seasonal travel, and when interstate travel restrictions are lifted, and then travel restrictions between countries are lifted, that's another change.

As hard as I try to understand the concept of herd immunity, and how populations become less susceptible to a specific virus over time, I still wonder if we won't see a pattern of one area doing better, then having another small outbreak of new cases, then doing better again, up and down. Maybe we need to prepare ourselves for that expectation (?)

2 Likes Save    
Thank you for reporting this comment. Undo
Ann

Another new release of the model today. I've been disappointed in the big changes to this model so far, so I'm not following it closely as I once was. It seems like today's changes are fewer.

But, for those who are still following and interested, there is a new release you might choose to check out.

1 Like Save    
Thank you for reporting this comment. Undo
huachuma

In our county in the Central Valley of CA we've had a few more reported infections but only one additional death since Tuesday. Keep isolating!

Save    
Thank you for reporting this comment. Undo
Moxie(Z4 St. Paul, MN)

Models always interest me because I love math and science. This one started out pretty rough, which is to be expected when trying to model something new and more complex than widgets coming off an assembly line. It has improved, in part, because there is more historical data to work with. Aside for any issues with the modeling assumptions and algorithms, the data they work with is inconsistent. For example, there is more than a little variability from state to state on how cases and deaths attributed to COVID-19 are counted. MN's department of health has steadily improved it's website and transparency about data. If the model uses that historic data, it might explain why the model is doing a much better job of predicting the trend here.

Since one of the assumptions of the model is that there are "stay at home" orders and that people abide by them, human behavior to the contrary can easily break the model.

3 Likes Save    
Thank you for reporting this comment. Undo
mudhouse

The IHME model in this thread cut the predicted total of deaths (thru Aug 4) in NM in half today. Now, they predict under 200 total deaths. I track our daily NM stats against the IHME model, and it's matching the course of our daily increases really well. Good, that's kind of reassuring.

But, yesterday, my NM governor held a presser saying our state is doing their own modeling (no source stated.) They've rejected the IHME model we've been discussing in this thread. They predict it will be off from my state's real numbers by a factor of 8 to 10 (!!!)

My state government says we'll peak in late May, not April 24, so a month later than IHME.

My state government said yesterday that we need 1821 ICU beds at peak. Today's IHME model shows we need 62 ICU beds at peak.

I think my state government has gone mad, but what do I know? I'm hoping eventually they'll release the source of their data and whose model they're using. At this point, I'm trying to rely on my sense of humor, and wine, to get through this.

2 Likes Save    
Thank you for reporting this comment. Undo
lurker111

They moved our peak date again. That model is wishful thinking. I wouldn't trust anything it says. I've seen better models from high school students. (not covid models)

1 Like Save    
Thank you for reporting this comment. Undo
maifleur03

mudhouse given that both your state and mine have seen relatively few cases and our states are away from any major outbreak area even thinking that the peak is in a week or two has never seemed realistic to me. Think of the virus as if it was an incoming tide. Tides take a while to cover areas except for bore tides they do not come at a rush but as an encroachment with not all areas being hit at the same time.

Edited to add that you know that there are many people on the various reservations in New Mexico who lack health care. Not certain given the quasi-citizenship status if their deaths, if reported, would be included in the general death count.

Save    
Thank you for reporting this comment. Undo
lurker111

Our latest update...


1 Like Save    
Thank you for reporting this comment. Undo
HU-5213567

ann - perhaps your state leadership is planning for the worse and hoping for the best. in a prior life, I was the charge nurse in a busy urban emergency department for 8 years. I have two ear piercings that i got along with my teammates (a few got a tattoo) after two night shifts where we were literally thankful we didn't find dead bodies in the waiting room. i cannot imagine consecutive shifts like that and hopefully your state leadership is doing everything they can to prepare so your healthcare providers are not overwhelmed.

eta - should say mudhouse. sorry about that.

2 Likes Save    
Thank you for reporting this comment. Undo
lurker111

The college is closed so the kids are gone.

1 Like Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

They've rejected the IHME model we've been discussing in this thread.

I don't believe California is using that model either. May has been predicted as our peak month, not the specific date of April 13 like the IHME model. Governor Newsom is still speaking of locating supplies and adding hospital beds, the opposite of what IHME predicts.

Los Angeles safer-at-home orders have been extended through May 15.

Los Angeles County cases are still increasing, as are cases in most all the counties across the state, but at different rates.


Save    
Thank you for reporting this comment. Undo
mudhouse

Maifleur, for some reason the controversy over peak time doesn't bother me personally; I'm in for the duration, and retired. (I do think it must be horribly hard for businesses to have so much uncertainty.)

But my state disagreeing with a national model about the number of deaths and medical resources by a factor of ten, after some states have already peaked and we have some solid US data, makes me a little bit...cranky.

Thanks for that info about California not using IHME, Nancy. It's mostly the factor of ten part, the degree of disparity, that bothers me.

My state is still increasing too, and in vastly different rates across the counties. The northern counties with Native American reservations and pueblos seem to be showing the highest per capita number of cases. (Maybe that's something being missed by IHME, since the Native population looks to me to be highly at risk.) Fortunately 90% of my state's ICU beds are in the northern part of the state, too.

My own county in the southern part of the state is close to Lurker's stats above. We only have 44 active cases and no deaths so far.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

Think of the virus as if it was an incoming tide. Tides take a while to
cover areas except for bore tides they do not come at a rush but as an
encroachment with not all areas being hit at the same time

Funny that you say that. A friend told me that he had a dream that he saw rising water in his neighborhood, then a tidal wave hit. He said he thinks it was a warning about Covid. We're 200 miles from the coast. I don't think anything about dreams, but he thinks they are messages from beyond.

Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

Which states are supposed to have already peaked?

ETA: Link to number on individual states - https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html

Save    
Thank you for reporting this comment. Undo
maifleur03

Be cranky all you want. However look at the map of the US and see where most of the people currently with the virus are. If people keep wanting and using their constitutional rights as free citizens and simply ignore the warning as I am seeing in my state the number of deaths predicted will be a lot less than what finally happens.

Save    
Thank you for reporting this comment. Undo
maifleur03

lurker while I do not believe in dreams as such I do believe in precognition because it has happened too many times in my life.

Save    
Thank you for reporting this comment. Undo
mudhouse

Which states are supposed to have already peaked?

Off the top of my head, according to IMHE, Louisiana, New Jersey, and New York have already hit their peak resource needs. I'm not sure if there are others, those are the ones I was thinking about.

Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

Rather than depend on a few day's worth of data, much longer is needed to determine a descending trend.

Italy stayed at a plateau for over a week before slight decreases in daily cases were noted. They're still registering 4,000 new cases in a day.

Peak only means the point of the maximum number of cases, not that there are no more new cases. New cases mean the contagion is still a threat.

Louisiana seems to have very recently had a slowing - https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html

1 Like Save    
Thank you for reporting this comment. Undo
mudhouse

I tried to access that link, Nancy, but I don't subscribe to the NY Times. I tried to get around the paywall by using archive.is, and I can, but then the page isn't interactive for me.

Peak only means the point of the maximum number of cases, not that there
are no more new cases. New cases mean the contagion is still a threat.

Yes, I know. I'm just saying "peak" indicates a substantial amount of information has been collected for at least some US states, and that should be more helpful to US prediction models than data from European countries. We should be getting better info, gradually.

IHME uses known deaths for their data. When it launched on March 26, the only country that had peaked was Wuhan China. So they used Wuhan.

After cities in Spain and Italy peaked, they were able to refine the model greatly because they had the death data from those countries as well. They use the time from the first social distancing measures imposed to the peak death rate, and apply that information to models for US states.

From my reading of IMHE's explanations of updates, the accuracy of the model will increase for later states as they're able to collect the death data from the earlier-peaking states.

That's what I was referring to. It's my understanding that all models will be improving their accuracy as they gather more real time data.

When my state announced their own (very disparate) figures yesterday, I felt like I was moving away from having a bit of clarity, not towards it. That was my honest frustration, whether it was justified or not. (I'll leave it to the mathematicians to argue out, I guess.)

Save    
Thank you for reporting this comment. Undo
Ann

"ann - perhaps your state leadership is planning for the worse and hoping for the best. in a prior life, I was the charge nurse in a busy urban emergency department for 8 years. I have two ear piercings that i got along with my teammates (a few got a tattoo) after two night shifts where we were literally thankful we didn't find dead bodies in the waiting room. i cannot imagine consecutive shifts like that and hopefully your state leadership is doing everything they can to prepare so your healthcare providers are not overwhelmed."

Zeus, did you mean to address this to mudhouse?

1 Like Save     Thanked by HU-5213567
Thank you for reporting this comment. Undo
mudhouse

That's what I thought, Ann.

1 Like Save     Thanked by HU-5213567
Thank you for reporting this comment. Undo
HU-5213567

Ann

"ann ...

Zeus, did you mean to address this to mudhouse?

_____

yep. sorry about that.

3 Likes Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

After cities in Spain and Italy peaked, they were able to refine the model greatly because they had the death data from those countries as well.

This is what worries me. From reading the Italian press, there is not agreement among their experts when the descent began/is beginning. Yesterday headlines were saying the descent in new cases was starting on the national level. Statistics that I have seen from Italy are on regions -- the equivalent of our states -- and by province -- the equivalent of our counties. Nothing on cities.

From when the country went into lockdown on March 9 until April 9 is a lot longer than any timeline we have in the US.


Link to data on the region of Lombardy (Lombardia): https://lab.gedidigital.it/gedi-visual/2020/coronavirus-i-contagi-in-italia/lombardia.php?refresh_ce

After a month of shelter in place, there is still a .07% increase in the number of persons hospitalized with coronavirus symptoms (ricoverati con sintomi), and a 2.5% increase in the number of positive cases isolated in their homes (isolamento domiciliare). Persons discharged or recovered is at 3.7% increase (dimessi/guariti), and deaths increased by 2.2% (decessi).

From the above data, I don't have a lot of faith in an earlier rather than later date for easing of restricted movement in the US.

Save    
Thank you for reporting this comment. Undo
mudhouse

From the above data, I don't have a lot of faith in an earlier rather than later date for easing of restricted movement in the US.

I think we're talking about apples and oranges a bit, Nancy, and maybe it's my fault for being unclear. I'll take one more run at it.

Although I listed the change in date, I went on to say it wasn't the change in peak date that bothered me, it was the degree of disparity in the number of deaths, and the amount of medical resources required; the fact that my state says the IHME numbers will be off from the reality of my state's numbers by a factor of ten. This large disparity just seems odd to me, since we're now at a point when the modelers know the actual peak deaths per day, and peak medical resources used, for a few states.

Instead, I think you're talking about the amount of time required to come down from the peak to reach a level of readiness to ease restrictions. Valid topic (and a hard one, for the reasons you mentioned, and more) but as I said above, that's not really what frustrates me. Instead I'm talking about disparate projections for the number of deaths and beds.

IHME now projects that NM needs 62 ICU beds at peak, and my state's model (according to the presser) projects we need 1821 ICU beds at peak.

Slight difference, eh?

IHME says projects NM's total deaths (through Aug 4) at 194, and my state's estimate, if IHME is "off by a factor of 8 to 10," is 1552 to 1940 deaths.


****************************

Anyway, if you're curious about how IHME dealt with locations in Spain and Italy, that page is here (I meant to link it above for you, and forgot.)

http://www.healthdata.org/covid/updates

Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

Lombardy still has volunteer foreign doctors to meet the demand in their hospitals. And they are still using field hospitals.

As I wrote earlier, California is using a model other than IHME, and I trust the state government and their experts to choose the best model for our anticipated needs. California's excess supplies can be sent to other Covid hot spots in the country.

Save    
Thank you for reporting this comment. Undo
mudhouse

Nancy, just curious, have you read anything about how Italy classifies deaths for Covid19? Many weeks ago, I read an article that had a statement from a doctor in Italy, who said in his hospital, they attributed any death in the hospital during the Covid19 epidemic as being caused by Covid19. (Sorry, too long ago, no link.) I was really surprised, and it was the first time I considered the problems caused by different countries possibly using different methods for classification of the cause of death.

I've just been wondering if you were familiar (by chance) with Italy's system.

Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

how Italy classifies deaths for Covid19?

I think I read something similar to what you did.

There have been some discrepancies found between records of the number of deaths reported -- between hospitals and the local government?

There's also a question of consistency in reporting deaths at home -- Covid or ?

Lombardy's hospitals were overwhelmed by the virus, and they were functioning in the proverbial fog of war.

Over 100 doctors and nurses have died from Covid.

1 Like Save    
Thank you for reporting this comment. Undo
Ann

Another new release of the IHME model is out today. Now, my state is supposedly 16 days past its peak, with our peak having been on March 28th. I'll be curious to see what our Dem governor does about opening our state. That's become quite the topic among governors - at least the Dem governors in recent days. I haven't heard much from the GOP governors quite yet.

ETA: Our expected death number hasn't changed much since the last update. It was 2400 on the 1st, then moved into the 200s and has settled in the 400s for two updates in a row now. Our peak date has continued to move backwards.

Save    
Thank you for reporting this comment. Undo
Kathy

Good news for your state. Hopefully people stay mindful and it doesn’t relapse. A new normal will most likely be the reality for the at least the near future.

Save    
Thank you for reporting this comment. Undo
foodonastump

Sorry to see the US total bump back up from 60,415 to 68,841. Wrong direction, folks, stay vigilant!

2 Likes Save    
Thank you for reporting this comment. Undo
Ann

Kathy, I'll use a new normal in terms of social distancing, wearing a face mask at the grocery store, and tons of hand washing - but I'll be very disappointed if my state doesn't begin its opening process on the 1st of May. I'm not a fan of our governor and it will be his decision. We'll see what he decides. My state has the best economy of any state in the nation, so I think that along with our early peak date presents an opportunity to put our state back to some level of normal soon, but we'll see what Polis decides.

Save    
Thank you for reporting this comment. Undo
HU-5213567


sigh..


Save    
Thank you for reporting this comment. Undo
Ann

FOAS, I thought the 60,415 was two updates ago. Maybe I'm wrong. I've quit writing it down, but I thought I remembered a 63,000 something more recently. I could probably review this thread to find out, but too lazy right now.

I do wish we could just temporarily remove New York from the numbers, though. It is amazing to me that 1 of 50 states accounts for nearly half the deaths in the nation, but that's the way it played out.

ETA: I just checked worldometer and New York accounts for nearly half our deaths at present, but is supposed to end up being about a quarter of the projected deaths when it's done (according to the IHME modeling).

Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

Today was supposed to be California's peak before IHME pushed the date of peak resource use back to April 17.

Save    
Thank you for reporting this comment. Undo
ubro(2a)

Our province of Sask. has now more recovered Covid in a day than active cases. We are doing not bad so far. So I cautiously hope our peak is past.

I attribute that in part to being later hit than other provinces so when the federal Quarantines went into place we had very few cases. Our provincial govt. also put strict penalties for anyone breaking their 14 day self isolation, among the early closure of non essential businesses, schools etc.

That said, I don't think our success, so far, is because we are better at handling this or at obeying orders. IMO it is partly because we are cold! No one uses us as a vacation destination in the winter. Luckily the last 7 days have been stormy, freezing and downright nasty. It is not hard to stay at home, cuddle up, read a book, watch a movie and eat in such weather.

1 Like Save    
Thank you for reporting this comment. Undo
foodonastump

Yeah, Ann, maybe NY shouldn’t be included. My own county already has far more deaths than your whole state is projected to have by August 4. Skews the numbers.

Save    
Thank you for reporting this comment. Undo
Ann

FOAS, Birx had to remove New York and New Jersey from her charts today in the news conference to even discuss the other metropolitan areas of the U.S. She said they skew the numbers so badly that we can't even see the lines of any other big city unless and until we remove those two states. To demonstrate her point, she had one chart with those two states and they were alone, with every other state at virtually zero. But, she then displayed a second chart where those two states were removed and we got a good picture of the other metropolitan areas.

It definitely caught my attention because that news conference happened not long after I was discussing that very topic here, so it was interesting to see it all in chart form.

1 Like Save    
Thank you for reporting this comment. Undo
bragu_DSM 5

chasing these numbers is like betting on the stock market ... but as more data becomes available, and more weight is given to social distancing, the more accurate they become, not that they will be accurate, because they don't have the algorithm figured out. Remember when the IMHE was predicting 250k deaths by Aug. 4?


now 114k deaths world round ...

Save    
Thank you for reporting this comment. Undo
Ann

The most recent IHME releases have certainly provided more consistency than 2-3 weeks ago. The April 1st release was quite a shock in its percentage changes. I guess more data has helped narrow the prediction changes - because they were almost absurd.

Save    
Thank you for reporting this comment. Undo
mudhouse

Today's update page has a lot of explanations: http://www.healthdata.org/covid/updates

Three days ago US deaths went from 60415 to 61545, now bumped up to 68841. The increase is partly explained by increased projections for Massachusetts and New York.

They say these states may be at or past their peak for hospital use: New York, New Jersey, Michigan, Illinois, and Louisiana.

These states may be nearing or reaching their peak for deaths this week: Ohio, Maine, North Carolina, Tennessee, Pennsylvania, and West Virginia.

They say these states may end up with the highest cumulative death tolls: Massachusetts, New York, Connecticut, Florida, New Jersey, Texas, and Georgia.

Save    
Thank you for reporting this comment. Undo
Kathy

No one state can open without the cooperation of neighboring states. That’s why governors are forming their own groups. Food transport and workers crossing states and even travel will have an impact on the rural areas maybe even more than the cities which have been the hot spots so far. Once people start traveling they will be carrying the virus to the more rural areas.

Save    
Thank you for reporting this comment. Undo
mudhouse

The good news is the IHME model decreased the projection for the UK, from 37494 down to 23791.

But they've increased Sweden's projection from 13259 to 18322. The last article I read said that Sweden is still trying to weather the pandemic using herd immunity, instead of shutting things down.

The projections for Italy, Spain, and France went up too, but not by as large a number as Sweden.

1 Like Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

The projections for Italy, Spain, and France went up too

Italy is seeing a rather slow descent in the number of new cases.

While there are fewer patients in intensive care, and in hospitals in general, the number of persons who are positive and quarantined at home is rising.

I would guess that the expansion of testing -- as a prelude to the gradual lifting of movement restrictions in Italy -- is revealing more community contagion than previously suspected. Stay-at-home orders have been extended to May 3, but some stores have been allowed to re-open -- children's clothing, book, and stationery stores.

3 Likes Save    
Thank you for reporting this comment. Undo
lurker111

We had the first death in our county yesterday. A woman in her 80's died at home. I don't think she tried to get help because we didn't have anyone in their 80's on the list until now. I think she was tested after she died. They said she was living with relatives.

Save    
Thank you for reporting this comment. Undo
HU-5213567

back down.



2 Likes Save    
Thank you for reporting this comment. Undo
maifleur03

The new information provides when different states can relax social distancing except then it mentions with testing, follow up tracing, isolation, and limited gathering. Not much different than what is supposed to be followed now except with dates.

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

Like I said, they'll get it right after it happens. They said we peaked 3 days ago, now.

Save    
Thank you for reporting this comment. Undo
mudhouse

IHME moved up the peak date of all of the states I'm watching for family members. The death numbers they're projecting for Alabama still seem a little higher than is actually playing out, and the projected death count for New Mexico are a little lower than is actually playing out, but both are within their estimated range.

Two pressers ago, my state said NM's real numbers would be 8-10 higher than this model. Last presser they said IHME doesn't take our diversity or "rural frontier nature" into account. No updates about whether they still think we'll peak in late May. IHME says our resource need peak is today.

So IHME tries to give me a total number and a peak date. My state gives me a list of virtual museum tours to watch while I wait for another six weeks.

I'll take the IHME model. At least they try to explain their projections, and they seem to be tracking with the reality of our daily numbers.

(Lurker, my county has had one death so far, hope yours is still only at one.)

Save    
Thank you for reporting this comment. Undo
lurker111

Yes, still just one with 4 hospitalized in the county. This is for my city...

Confirmed Cases: 36
Active Cases: 19
Fatalities: 0
Recovered: 17

1 Like Save    
Thank you for reporting this comment. Undo
lurker111

I've heard that we've had a 2nd death, but haven't seen it documented.

"Someone very young" is the rumor. We have 2 documented cases of children in the 0-9 year range.

Save    
Thank you for reporting this comment. Undo
Moxie(Z4 St. Paul, MN)

Some assumptions in the IHME model are extremely optimistic. The model only addresses a short timeframe, which is problematic because it is highly unlikely that COVID-19 will be erradicated in that timeframe. On the plus side, they do give a wide range of values for their predictions even though most people probably just look at the prediction lines.

Their estimates for MN went well at first, but their predicted values for deaths is optimistic now. The death rate is still within the predicted range, however.

1 Like Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

Caution Warranted: Using the Institute for Health Metrics and Evaluation Model for Predicting the Course of the COVID-19 Pandemic

https://annals.org/aim/fullarticle/2764774/caution-warranted-using-institute-health-metrics-evaluation-model-predicting-course

The IHME projections are based not on transmission dynamics but on a statistical model with no epidemiologic basis. Specifically, the model used reported worldwide COVID-19 deaths and extrapolated similar patterns in mortality growth curves to forecast expected deaths. The technique uses mortality data, which are generally more reliable than testing-dependent confirmed case counts. Outputs suggest precise estimates (albeit with uncertainty bounds) for all regions until the epidemic ends. This appearance of certainty is seductive when the world is desperate to know what lies ahead. However, the underlying data and statistical model must be interpreted cautiously. Here, we raise concerns about the validity and usefulness of the projections for policymakers.

4 Likes Save    
Thank you for reporting this comment. Undo
Blanche Dubois

Thank you for this article Nancy

Save    
Thank you for reporting this comment. Undo
Ann

Zeus, I'm glad you posted about the new update. I saw the new IHME update and wanted to post about it, but I couldn't locate the thread anymore - so I'm glad your update bumped it back up.

Save    
Thank you for reporting this comment. Undo
maifleur03

I find it very interesting that states with major metropolitan areas can vary so widely in the estimated relax dates. However I did notice those dates are followed by the use of tracing which few states are bothering to do. I went to get some milk and other things this morning. There is no way that I could tell you who the clerk at the store I purchased the milk from is nor the person who passed me going in. I also noticed some states the peak has moved forward. Iowa, Nebraska, North and South Dakota are now late June.

Save    
Thank you for reporting this comment. Undo
lurker111


Save    
Thank you for reporting this comment. Undo
Ziemia(6a)

The Corps has greatly expanded the number of beds the last 3 weeks.

Save    
Thank you for reporting this comment. Undo
heri_cles

Thank you Nancy.

Save    
Thank you for reporting this comment. Undo
HU-5213567

up a bit.


Save    
Thank you for reporting this comment. Undo
HU-5213567

sigh...


Save    
Thank you for reporting this comment. Undo
lurker111

It's all within the huge margin of error. Can't go wrong.

Save    
Thank you for reporting this comment. Undo
HU-5213567

I wish I could say I was surprised.


Save    
Thank you for reporting this comment. Undo
maifleur03

I am expecting deaths to go over 100,000 now that the states are easing the stay at homes. My governor even stated that there was no need to maintain the 6 feet if you are working.

Save    
Thank you for reporting this comment. Undo
dandyfopp



Save    
Thank you for reporting this comment. Undo
HU-5213567

down a bit


Save    
Thank you for reporting this comment. Undo
Ziemia(6a)

Weird cuz we're averaging over 2k deaths daily and are over 61k so far. We'll reach 72k in maybe less than a week.

Save    
Thank you for reporting this comment. Undo
lurker111

You're wasting your time with that junk website.

Save    
Thank you for reporting this comment. Undo
HU-5213567

^^ yes. the model assumes distancing recommendations. we all know that is getting ignored by many now and more soon.

Save    
Thank you for reporting this comment. Undo
HU-5213567

lurker111

You're wasting your time with that junk website.

————

you are wasting your time on this thread.

Save    
Thank you for reporting this comment. Undo
HU-5213567

very sad


Save    
Thank you for reporting this comment. Undo
maifleur03

I noticed several changes on this even today. When I looked earlier at Missouri there were no projections of deaths from today onward. But that has been updated. It has added testing as a separate grouping which is great if there is actual testing materials available to use. The city was given 500 tests to use and the time slots were filled in under an hour.

Save    
Thank you for reporting this comment. Undo
Ziemia(6a)

They recently doubled it?

Weren't there comment celebrating the 60k numbers & thanking Trump for them?

Save    
Thank you for reporting this comment. Undo
nancy_in_venice_ca Sunset 24 z10

I noticed for California when the number of daily deaths rose past the model's maximum-day-for-deaths that predictive information simply disappeared.

Save    
Thank you for reporting this comment. Undo
Moxie(Z4 St. Paul, MN)

IHME released a new version of the model on 4 May. Here's a link to description of changes, which are major. Some of their initial assumptions about the continuation of social distancing were optimist, so the predictions were often optimistic. I expect the use of mobility data by state will be a bit more realistic. There are still some baseline problems. For example, IHME doesn't have accurated data about how many ICUs bed are available in MN.

What Nancy described is a software bug, but I don't have time right now to dig around to consider what kind.

Save    
Thank you for reporting this comment. Undo
HU-5213567

sigh...



Save    
Thank you for reporting this comment. Undo
HU-5213567

so disappointed in us


1 Like Save    
Thank you for reporting this comment. Undo
studio10001

Me, too.

Save    
Thank you for reporting this comment. Undo
HU-5213567

not worse anyway


Save    
Thank you for reporting this comment. Undo
how2girl

Going very well...

9 deaths in total, currently 4 in hospital, 1 in ICU

Had 8 consecutive days with no new cases.

Total population approx. 2.7m


2 Likes Save    
Thank you for reporting this comment. Undo
maifleur03

This state Missouri seems to have semi-stabilized at between 11 and 16 deaths a day. Now up, as of this morning at 605 deaths. There are others but I do not think the state is counting those that are virus related.

Save    
Thank you for reporting this comment. Undo
HU-5213567

better....


Save    
Browse Gardening and Landscaping Stories on Houzz See all Stories
Modern Homes Sneak a Peek at 7 Homes From Denver’s Modern Home Tour
A wine cube with a glass ceiling, a remodeled Eichler and other structures exemplify modern design in Colorado
Full Story
My Houzz 19 Bathrooms That Aren’t Afraid of Color
Bold hues and pretty patterns add a splash of fun to these personalized homes
Full Story
Most Popular CES 2015: Inching Toward a Smarter Home
Companies are betting big on connected devices in 2015. Here’s a look at what’s to come
Full Story