Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
×
United States Politics

As States Move to Reopen, 2 Projections Show Deaths Rising (nytimes.com) 445

A University of Washington forecast frequently cited by the White House projected a death toll of more than double what it was predicting last month, citing the "easing of social distancing measures" in many states. From a report: As President Trump presses for states to reopen their economies, his administration is privately projecting a steady rise in the number of coronavirus cases and deaths over the next several weeks. The daily death toll will reach about 3,000 on June 1, according to an internal document obtained by The New York Times, nearly double the current number of about 1,750. The projections, based on government modeling pulled together in chart form by the Federal Emergency Management Agency, forecast about 200,000 new cases each day by the end of the month, up from about 25,000 cases a day currently. The numbers underscore a sobering reality: While the United States has been hunkered down for the past seven weeks, significant risks remain. And reopening the economy will make matters worse. "There remains a large number of counties whose burden continues to grow," the Centers for Disease Control and Prevention warned.

As the administration privately predicted a sharp increase in deaths, a public model that has been frequently cited by the White House revised its own estimates and projected a death toll of more than double what it was predicting last month. The Institute for Health Metrics and Evaluation at the University of Washington is now estimating that there will be nearly 135,000 deaths in the U.S. through the beginning of August -- more than double what it forecast on April 17, when it estimated 60,308 deaths by Aug. 4. (There have already been more than 68,000 deaths in the U.S.) The institute wrote that the revisions "reflect rising mobility in most U.S. states as well as the easing of social distancing measures expected in 31 states by May 11, indicating that growing contacts among people will promote transmission of the coronavirus."

This discussion has been archived. No new comments can be posted.

As States Move to Reopen, 2 Projections Show Deaths Rising

Comments Filter:
  • by Anonymous Coward

    so here's some NY Times hype about model predictions. When these are also wrong they'll have new, more dramatic ones next week.

    • by MightyMartian ( 840721 ) on Monday May 04, 2020 @06:17PM (#60022216) Journal

      Um, model predictions have been wrong because most developed countries have put pretty radical physical distancing measures in place. Reducing the attack surface for the virus means the model those measures were based on wrong. But that demonstrates the success of the model; in that it informed various governments as to the best way to avoid the projections the model made.

      This is like some flood planner developing a model that a river, if left to its own devices, will flood a town, and local government building a big dam or dike to prevent the flood waters from reaching the town, and then facing down some asshole on the Internet who says "You see, the town didn't get flooded, so you're just an alarmist!"

      The one thing I would expect on /. would be people who at least have some passing familiarity with how mathematical models are formulated, but perhaps the biggest casualty of COVID-19 has been peoples' wits.

      • by kbahey ( 102895 ) on Monday May 04, 2020 @07:18PM (#60022514) Homepage

        Um, model predictions have been wrong because most developed countries have put pretty radical physical distancing measures in place. Reducing the attack surface for the virus means the model those measures were based on wrong. But that demonstrates the success of the model; in that it informed various governments as to the best way to avoid the projections the model made.

        This is like some flood planner developing a model that a river, if left to its own devices, will flood a town, and local government building a big dam or dike to prevent the flood waters from reaching the town, and then facing down some asshole on the Internet who says "You see, the town didn't get flooded, so you're just an alarmist!"

        The one thing I would expect on /. would be people who at least have some passing familiarity with how mathematical models are formulated, but perhaps the biggest casualty of COVID-19 has been peoples' wits.

        Well said.

        Even a more relevant example (for /. audience), is Y2K: the prediction of doom and gloom did not materialize precisely because the risk was mitigated by lots of time, effort and money poured into fixing the problem before the hard date.

        Despite that, you still see some drooling idiots who spout things like: "Oh, it was overhyped, nothing happened".

        Just a week ago, I had a discussion with such an idiot right here on /., and he was saying things like : "don't change the rules mid game". No! You do change the rules because all governments are adapting to whatever the situation on the ground happens to be, based on what the new data says, not something from month ago.

        An example here in Ontario (and Quebec): the feared scenario of clogged intensive care units and ventilator shortages did not materialize. And that is good because society was on lock down. However, outbreaks spread through retirement and long care homes like wildfire. Do we still buy ventilators because that was the rule of the game, or instead we mobilize resources for where they are needed (e.g. protective equipment and army mobilized to care homes)?

        And 3 months from now, it will be different, and later this year it will still be more different ... and so on. That is just good crisis management, not changing the goalposts of the game ... Only a rigid (the inflexible conservative?) mind would complain about that ...

    • by JoshuaZ ( 1134087 ) on Monday May 04, 2020 @06:28PM (#60022280) Homepage
      The model in question is the IHME model, and it definitely hasn't been terribly accurate. Until this update it was predicting 57,000 deaths as the low end by May 23rd, when we had already surpassed that while it was predicting that. But models are by nature noisy, and very small changes can have large impacts. A better approach then is to look not at a single model, but a collection of major models. This https://projects.fivethirtyeight.com/covid-forecasts/ [fivethirtyeight.com] has six different models, (one of which is IHME) so you can compare them, and discusses also what each model is predicting.
      • by RhettLivingston ( 544140 ) on Monday May 04, 2020 @08:40PM (#60022682) Journal

        Given that the model does not model the response to the model, it cannot be accurate when looked at in hindsight. It is always predicting a future that will never happen.

        Every release is predicting what will happen if nothing changes in the plans to get this under control or in people's habits vs. the point in time at which it is released. But changes are made.

        We look at the model's prediction, see that we don't like it, do more, successfully change our outcome, and make the model wrong. That is exactly the goal of releasing the model!!!!

  • by Rick Schumann ( 4662797 ) on Monday May 04, 2020 @06:30PM (#60022288) Journal
    Some things I'd like to point out, some (or all) of which may not be popular (and will likely draw lots of fire):
    0. Yes, the pandemic is a Real Thing, it's not a hoax, it's not Amateur Night-level shit, it's not 'the flu', or any nonsense like that. Disclaimer over.
    1. Pick your poison, America: We can keep everything shut down for the next year or two, and for all we know there might be no more deaths from the virus, or they'll be kept to a minimum -- but there wouldn't be a United States left, because the entire country would have collapsed economically to the point that the Great Depression would look like a minor blip comparatively speaking.
    2. Assuming the worst-case scenario from the above: there might actually end up being more deaths anyway, because with a completely collapsed economy hospitals and doctors would run out of supplies and medicine, making treatment almost impossible. Might not even have electricity for that matter. Basically, the United States becomes a third-world country.
    3. There has to be a middle ground. Therefore things must start back up again -- in a slow, careful, measured way. 4. You can't fix stupid. Even with all the current restrictions, some people are still managing to be phenomenally stupid. Call it 'evolution in action', I guess. So when things start opening back up, is there going to be some increase in infections and/or deaths? Yes. Considering all I've said above, can we really avoid this? No, I don't think so. Again, 'call it evolution in action', I guess.
    5. So, America, and Slashdotters: Hold on to your hats, take your Xanax, and Don't Be Stupid or be around Stupid People, because the United States is coming out of Standby whether you personally agree with that or not.
    • by NFN_NLN ( 633283 )

      > Some things I'd like to point out, some (or all) of which may not be popular (and will likely draw lots of fire):
      > 0. Yes, the pandemic is a Real Thing, it's not a hoax, it's not Amateur Night-level shit, it's not 'the ***', or any nonsense like that. Disclaimer over.

      Can you qualify that for (a) 60 years of age?

      For anyone 60 years of age and in good health, this doesn't appear to be any less dangerous than a typical "virus" whose name I can't say least ye be triggered.

    • I'm not running for office, but I approve of this comment. People exercising normal freedoms will lead to a higher death rate anyway, e.g. more driving, etc. It's the age old question of at what point is a life so well protected that it is no longer worth living.

    • And the middle ground is what a month or two of social distancing is trying to accomplish. Control the infection rate so it remains manageable, then start reopening businesses, but maintaining social distancing measures. The trick is to find the balance point where you get the economy going again, while trying to stave off a second wave. For a lack of a better word, health officials hope to "train" people to continue the social distancing while businesses are reopened, because the next ugly period is this f

      • If you favor extending the lock down, you're just favoring other demographics as expendable such as those with cancer, heart and other major problems. Already happening. Then include the wrecked futures of perfectly healthy people. The middle ground is over.
    • by HiThere ( 15173 ) <charleshixsn@ear ... .net minus punct> on Monday May 04, 2020 @07:18PM (#60022512)

      You're missing several choices that have just been ignored.

      Contact tracing of cases if very important, and if that is in place, then a lot of restrictions can be lifted. But it requires a lot of "feet on the ground". Nobody in the US seems to be implementing it, though there are some recent stories that indicate that New York (I'm not sure whether the city or the state) is beginning to staff up for it.

      Accurate tests for both active cases and antibodies are extremely important. The current tests are mainly useful for getting population statistics, because they have both too many false positive and false negatives. And a lot of the ones used in the US don't even have validated figures for how many false positives/negatives they get, so they aren't even useful for population statistics. Without the accurate tests a lot of the active cases of COVID will be missed, so distancing becomes more important.

      • by AmiMoJo ( 196126 )

        Germany is going back to square one with contact tracing, starting from scratch. They are going to trace everyone from the very beginning. Seems like the opportunity is not still there if you are willing to put in the effort.

        Contact tracing is the only way we can get the economy moving again while we wait for a vaccine.

    • Here's a picture of what's possible. Australia has just 95 covid-19 deaths. That is not a typo. As far as we know, no deaths are being swept under the carpet (ie, attributed to other causes) because we have one of the highest rates of testing in the world [abc.net.au].

      We have almost entirely eradicated the virus [worldometers.info] through aggressive testing and contact tracing.

      All of this will cost us AUD$192bn (about USD$123) through all of the government stimulus payments. To raise that money, the government is issuing bonds at 0.4 [rba.gov.au]

  • Some number of people will die. How many? We don't know because the testing we're doing sucks. Between a, I'm probably miss-remembering, a 20% both false positive and 20% false negative test, we also have the fact that we're only testing people who want to be tested. These statistics are 100% bogus and you can't draw any conclusions from them. Unless your a politician or an activist that wants to score points.

    What we need is A) a reliable test that gives results in, I dunno, 4-5 days works for me.
    • The data coming back from the early days of the pandemic, at least in Europe and North America, suggests people were already dying from complications of COVID-19 even before we had a firm handle on how contagious the virus was. Extrapolating from that is tricky, but it does suggest that the virus arrived earlier (hence, even the most vigorous border-closers had already closed the barn door after the cows had come home), and that it is more widespread now. We do know from the rates of reported infections in

    • Such a study has been done [reuters.com], in the closed, contained environment of prisons.

      In one test, 2300 were tested, 87% were infected, and 95% of those infected were asymptomatic.

      In another test, nearly 4700 were tested, 70% were infected, and 96% of those who tested positive were asymptomatic.

      NONE of the models or statistics assumed that 95%+ would be asymptomatic. Perhaps we should have followed the Swedish model (which did not lock down, but said "just be careful" [nypost.com]) and not destroyed our economy and spent

      • by HiThere ( 15173 )

        Two points:
        1) Asymptomatic at that time. Many asymptomatic cases go on to become more serious cases.
        2) Do you really trust a prison doctor who says that someone is asymptomatic? Those folks have occasionally said that dying people were healthy.

  • They project 200K cases a day? So much for being over the curve. That's like "we haven't seen anything yet".
  • by Solandri ( 704621 ) on Monday May 04, 2020 @06:52PM (#60022388)

    University of Washington is now estimating that there will be nearly 135,000 deaths in the U.S. through the beginning of August -- more than double what it forecast on April 17, when it estimated 60,308 deaths by Aug. 4.

    The death rate (deaths per day) will increase if you open early. But if you do it right there will be no difference in total deaths by the time this is all over with.

    As you flatten the curve, the area under the curve remains the same. That is, flattening the curve lowers its height (number of people infected at any given time), but extends it in time (the virus outbreak lasts longer). The area under the curve (total number of people who catch the virus) remains the same.

    So if you compare the death projection through August 4 if we don't reopoen, to August 4 if we do reopen, of course the latter will have more accumulated deaths by August 4. But if you compare the projected deaths up until when the virus drops off the radar, then the cumulative number of deaths for both scenarios will be the same. As long as you don't let the peak rise high enough to overwhelm your hospital capacity, the number of deaths by the time this is all over will not change. That's the only part you need to get right when flattening the curve - don't let your hospitals get overwhelmed.

    It's like opening a squeaky window. If you open it too quickly, it'll hurt your ears (hospitals will get overwhelmed). If you open it really slowly, the squeaking at any given moment is less, but it'll take a lot longer to open the window so the total amount of squeaking is the same as if you open it at moderate speed. So it's better to open it at a moderate speed - as quickly as you can without hurting your ears. That'll let you minimize the time you have to deal with opening the window (minimize the economic damage). Ideally we'd develop a vaccine (oil the window), but the experts pretty much say it'll be a miracle if one is developed and distributed within 12-18 months, and 4 years is more likely if we're even able to develop a vaccine. I don't think our economy can survive being shut down like this for 12+ months.

    • by cusco ( 717999 ) <brian.bixby@NOSpam.gmail.com> on Monday May 04, 2020 @07:25PM (#60022532)

      the cumulative number of deaths for both scenarios will be the same

      No, because an overloaded health system without sufficient supplies or equipment will cause more deaths which would be preventable if the consumption can be spread out over a longer time.

      If your local hospital has 20 respirators and they have 40 people who need them at one time then half your critical patients are going to die. If those 40 people get spread out over a longer time then as people are released respirators can be devoted to new patients. I really don't understand why this concept is so frelling hard for conservatives to grasp. Oh, that's right, Barbie says, "Math is hard!"

    • by Jesus H Rolle ( 4603733 ) on Tuesday May 05, 2020 @04:24AM (#60023578)

      The death rate (deaths per day) will increase if you open early. But if you do it right there will be no difference in total deaths by the time this is all over with.

      As you flatten the curve, the area under the curve remains the same. That is, flattening the curve lowers its height (number of people infected at any given time), but extends it in time (the virus outbreak lasts longer). The area under the curve (total number of people who catch the virus) remains the same.

      So if you compare the death projection through August 4 if we don't reopoen, to August 4 if we do reopen, of course the latter will have more accumulated deaths by August 4. But if you compare the projected deaths up until when the virus drops off the radar, then the cumulative number of deaths for both scenarios will be the same. As long as you don't let the peak rise high enough to overwhelm your hospital capacity, the number of deaths by the time this is all over will not change. That's the only part you need to get right when flattening the curve - don't let your hospitals get overwhelmed.

      This ignores the possibility of a treatment that significantly reduces the mortality rate. If it's discovered that some previously untested drug *cough* nicotine *cough* keeps most people alive, it's best if we've kept the curve flat until that discovery. A vaccine isn't the only possible medical solution to this pandemic.

  • > As States Move to Reopen, 2 Projections Show Deaths Rising

    Wake me when projections show cumulative deaths DECREASING!

    'Cause then we got Zombies!

  • Compounding the lack of testing, there's also a total dearth of understanding on the actual mortality rate - do a little research into 'excess death' stats. Many areas have a HUGE (order of magnitude) of deaths recorded, but aren't confirmed COVID-19.

    If a city normally has 100 deaths in a week during this week of the year, but instead are recording 1,000 (but not attributing to COVID-19) - what you you think is actually happening?

    Until we understand this (and a number of other details), it's extremely prema

  • by wakeboarder ( 2695839 ) on Monday May 04, 2020 @09:37PM (#60022812)

    America can be opened and you don't have to participate (unless your still employed, then you have the option to quit).

    Fact #1: There is NO good way as of today to stop COVID 19, that means you mean and at least ~60% of America will get this disease at some point in time, because as of today a vaccine is 18 months away (unless something crazy happens).

    Fact #2: Staying at home will most likely only delay getting COVID because at some point, you will have to get on with life, you will have to face the risks. Flattening the curve doesn't mean you don't get sick, it means you don't get sick at the same time as everyone else.

"Just think, with VLSI we can have 100 ENIACS on a chip!" -- Alan Perlis

Working...