Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Government Medicine United States

Many Scientists Warn CDC's New Death Rate Estimates Far Too Low (buzzfeednews.com) 151

Slashdot reader DevNull127 quotes Buzzfeed News: Public health experts are accusing the CDC of bending under political pressure to say the coronavirus is less deadly. New CDC estimates of coronavirus death rates look suspiciously low and present almost no data to back them up, say public health experts who are concerned that the agency is buckling under political pressure to restart the economy...

While no one yet knows the coronavirus's actual death rate, the agency's range of possible rates seemed alarmingly low to many epidemiologists, compared to existing data in places both inside and outside the US. For instance, estimates of New York City's total death rate, 0.86% to 0.93%, are even higher than the CDC's worst-case scenario. Estimates from countries like Spain and Italy are also higher, ranging from 1.1% to 1.3%. Researchers also lambasted the CDC's lack of transparency about its data sources. The eight-page document disclosed almost nothing about its numbers, citing only internal data and a preprint — a study that has not been peer-reviewed — led by scientists in Iran. "This is terrible. This is way too optimistic," Andrew Noymer, an associate professor of population health at the University of California at Irvine, told BuzzFeed News, adding, "With this document, the CDC is determined to smash its credibility with the public health community of which it is supposedly a leader."

The CDC did not return multiple requests for comment...

A preliminary analysis of more than two dozen studies from Europe, China, the US, and elsewhere, conducted by Meyerowitz-Katz and colleague Lea Merone, suggests that the overall infection fatality rate is between 0.5% and 0.78%. Even the lower end of that range is higher than what the CDC says is its "best estimate" for the rate, which is about 0.26%. The CDC's proposed fatality rates "are more in line with a relatively mild seasonal flu season than with COVID-19," said Gerardo Chowell, a public health expert at Georgia State University. Those estimates are at least an order of magnitude lower than ones elsewhere in the world, he added, including South Korea, which has a case fatality rate around 0.7% and one of the highest testing rates for the coronavirus in the world....

The CDC document provided almost no sources for its projections, making it impossible for scientists to understand how it came up with them.

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

Many Scientists Warn CDC's New Death Rate Estimates Far Too Low

Comments Filter:
  • by Anonymous Coward

    ...by calling any death accompanied by even a suspected COVID infection, a COVID death

    • by Z00L00K ( 682162 ) on Saturday May 30, 2020 @09:45AM (#60125050) Homepage Journal

      Turn it around instead - if you look at the averages for the same period previous years and then the number of deaths this year, then almost all 'excess' deaths are likely Covid-19 since we haven't had any other major death factor around.

      • Re: (Score:2, Funny)

        by fuzznutz ( 789413 )
        People are dying now that never did before.
      • You are ignoring the increase in suicides and deaths that were preventable from things such as untreated heart attacks in people who are too afraid to go to the emergency room.
        • by ivano ( 584883 )
          But doesn't that get cancelled out by less car deaths, pedestrian deaths, pollution deaths etc We can play this game all day long. The virus is killing people and we have few ways of bringing that number down. Our instruments are blunt. We have to deal with that fact and not try and bully everyone to do the opposite of was the "experts" tell us because it makes us seem so, so much smarter.
  • Seems low (Score:4, Informative)

    by Geoffrey.landis ( 926948 ) on Saturday May 30, 2020 @09:46AM (#60125052) Homepage

    Right now, the US statistics are 102856 deaths and 1748705 confirmed cases, for a 5.9% death rate of CONFIRMED cases.
    (data: https://coronavirus.jhu.edu/ma... [jhu.edu] ).
    But the number of tests is limited, and hence the actual number of cases is much larger than the number of confirmed cases (and including the fact that 30-50% of the cases show no symptoms at all).
    The CDC's "best guess" estimate of 0.26% can only be accurate if for every confirmed case in the U.S., there are 22.6 actual cases.
    That seems a little high to me.

    • The numbers depend on a lot of factors. How much and where you test, for one. But also how you count: many deaths here went uncounted as Covid victims, so they started counting nr of total deaths as well, in excess of what is normal for the time of year. Breaking it down is relevant as well, I for one would want to know if for instance our high death count (in NL) might be due to the policiy of not issuing staff and residents in retirement homes - filled with the most vulnerable people - with any protect
      • by rtb61 ( 674572 )

        The numbers are pretty clear in Australia. Want better numbers, do what they did in Australia years ago, inoculate the population for TB, it seems to have worked really well, yet for some reason, everyone seems to want to ignore reality in this case. If you have been inoculated for TB than you a pretty safe, if you have not, then get that decades old inoculation for TB all indications are that it works well in preventing covid 19 from triggering pneumonia, the old fashioned name for SARS - 'suddenly acquire

        • It's not self-evident, it's an ongoing study. We're actually taking part in it.
        • Having every 10-14yr old between 1953 and 1988 given BCG in the UK doesn't seem to have stopped it having the worlds highest fatality/mil stats. A vaccine believed to protect for only up to 15 years here anyway.

          Like lots of things it is being tested as a treatment and prophylactic, speculative testing is no guarantee of anything.

    • Well, if there's one thing we've learned in this pandemic it's that getting the numbers to look good is the most important thing.

    • Re: (Score:3, Interesting)

      Up to around 80% [healthline.com], and in some studies 96% [reuters.com] of all CoVID-19 infected are asymptomatic. A 1:22 ratio doesn't really seem all that out of line.
      • Re: (Score:2, Interesting)

        by hankwang ( 413283 )

        NYC has 16k confirmed deaths and 200k confirmed cases on a population of 8M (0.2% and 2.5% of the population). That's really not consistent with 0.26% IFR and given that about 25% of the NYC population has antibodies. Based on antibody occurrence, NYC has 0.8% IFR.

        • Re:Seems low (Score:4, Insightful)

          by Mr D from 63 ( 3395377 ) on Saturday May 30, 2020 @11:48AM (#60125502)
          The IFR can go up or down based on the percentage of people infected in high risk demographics vs low risk demographics. If a country, region, or state is doing a better job of preventing infections in the at risk groups, their IFR will be much lower.

          I see all this arguing without accounting for simple and obvious factors. Nobody ever cares about data normalization, I think many don't even know what it means yat argue away with comparative statistics.
          • I see all this arguing without accounting for simple and obvious factors. Nobody ever cares about data normalization, I think many don't even know what it means yat argue away with comparative statistics.

            When looking at a blended population like a city in this case, ( or a country ) the data are already normalized due to the huge sample size. You don't need to restrict or normalize data for example for age ranges / nursing home occupancy unless the city or country is outside the mean of other cities or countries in the area being studied. When looking at overall statistics of a whole population you will not have to normalize for age ranges until you are looking at the statistics for those specific ranges. W

            • When looking at a blended population like a city in this case, ( or a country ) the data are already normalized due to the huge sample size.

              This is absolutely incorrect, perfectly demonstrating the lack of statistical insight in the general public I spoke of.

              • No, what I said is 100% correct. If you are looking at what affects a population you absolutely do NOT intentionally control for data variables that are already equalized across the population sample. That's called data manipulation. If we look for "deaths in the U.S. population" it would be disingenuous to inflate or deflate any ranges in the data. We are looking for the statistics on what happens to a blended population, so we can make the logical informed assumption that any minor outliers like a retirem

      • First, the CDC page we're discussing explicitly did not use those numbers. Their best estimate is in the article discussed; it was that 30% of cases were asymptomatic. The high end of their range was 50%. This is in line with most other studies.

        Second, the references you give are talking about people who are " contagious but not [yet] physically sick" -- this includes both asymptomatic cases, and also people who are infected but haven't yet developed symptoms.

      • by ceoyoyo ( 59147 )

        Sure. It would just require that extreme value (96%) is true, and also that testing almost never detects an asymptomatic case.

      • by quenda ( 644621 )

        Up to around 80% [healthline.com], and in some studies 96% [reuters.com] of all CoVID-19 infected are asymptomatic.

        No, you mis-read. Being unaware they have the virus is not the same as being asymptomatic. Lots of people are saying "it is just a cough and sniffle".

      • Up to around 80% [healthline.com], and in some studies 96% [reuters.com] of all CoVID-19 infected are asymptomatic.

        You've been pushing [slashdot.org] that [slashdot.org] lie [slashdot.org] hard for over a month now Lynnwood.

        Isn't it about time to get some new material?

    • This is a schizophrenic virus, or there really are two of them. Here in Washington State we currently have 21071 cases with 1111 deaths which is 5%. But Thurston county has 133 cases with only one death. And there are multiple counties that haven't had any deaths yet. The virus can't even find Garfield county. It definitely is not an open country bug.

      The other odd thing in Washington is that it is disproportionately killing white people. Hispanics are significantly under represented, possibly because the o

      • In my state, the failing rurall hospitals have no Covid-19. why? Because we sent all cases into the CITY to deal with them leaving whatever they had for healthcare free to deal with everything else. We even have taken in cases from neighboring states where our city hospitals were closer.

        This is skewing the stats towards the city further and the red necks into feeling safer than they are.

        • That is not the way they are counting here. The statistics list where you live, not where you end up in a hospital. The first death here in Grant County was listed as a Grant County death even though the woman died in the hospital in Chelan county. Obviously I can't speak for how they count them where you live.

    • I should modify that post slightly; there's a difference between the death rate of people infected, and the death rate of people who show symptoms.

      The "best estimate" on the CDC scenario page being discussed [cdc.gov] is that the death rate is 0.4% of the symptomatic cases, and 35% of all cases are asymptomatic, the total death rate of infected people is 0.3%.

      So, the difference between CDC numbers and the current numbers for the U.S. [jhu.edu] is not due to an assumption of the number of asymptomatic cases, since they explicit

      • by HiThere ( 15173 )

        The problem is that "show symptoms". It seems plausible that the primary action of COVID is on the walls of the blood vessels, and that this can lead in some cases to small circulating blood clots. These can produce all sorts of different symptoms, including many that are not usually associated with COVID. You don't hear much of strokes or kidney failure being associated with COVID, but those can be the only sign of it's presence that is visible without a test. There's a huge number of "rare diseases" t

  • by DogDude ( 805747 ) on Saturday May 30, 2020 @09:53AM (#60125078)
    I don't think there's any question that the CDC has been compromised by this administration. Just the other day (yesterday?), the pulled the suggestion that churches shouldn't hold choir practices because of the singing etc. They admitted that they pulled that recommendation because it wasn't vetted by the White House. I make COVID-19 decisions based on information from the WHO, not the CDC, as a result.
    • Don't wear a mask if you're healthy and not caring for a CoVID-19 patient. That's what the WHO says [fox6now.com].
      • by HiThere ( 15173 )

        I'm sorry, but your link doesn't appear to point to a WHO site. Perhaps you mistyped the link.

    • by znrt ( 2424692 )

      didn't you hear? china is lying!

  • Can't have Winnie-the Pooh make a temper tantrum, so report what he wants to hear, even if in this case it ain't a bear but an orange-utan.

  • TFA is wrong (Score:4, Informative)

    by Solandri ( 704621 ) on Saturday May 30, 2020 @09:57AM (#60125094)

    For instance, estimates of New York City's total death rate, 0.86% to 0.93%, are even higher than the CDC's worst-case scenario.

    The CDC's worst-case scenarios [cdc.gov] (scenarios 3 and 4) are a 1.0% fatality rate. Last I checked, 1.0% is higher than 0.93%, not the other way around. Scenarios 1 and 2 (best case) are 0.2%. And their "best estimate" based on data they're getting from the states is 0.4%.

    The problem here isn't the CDC bending to someone's political will. The problem is the media not fully reporting the facts, and bandying the 0.4% 'best estimate" rate as if it were final and absolute with zero uncertainty. Go to the link I gave above. See what the CDC is saying from the horse's mouth, not from someone in the media trying to make it look like the CDC is bowing to pressure. The CDC gave a fairly large range (0.2% - 1.0%) which is not inconsistent with all the other expert estimates I've seen.

    • For instance, estimates of New York City's total death rate, 0.86% to 0.93%, are even higher than the CDC's worst-case scenario.

      The CDC's worst-case scenarios [cdc.gov] (scenarios 3 and 4) are a 1.0% fatality rate.

      I'm not sure that these are not different things.

      You're quoting the CDC planning scenario's estimates of death rate of symptomatic cases. To get the total death rate, you need to multiply by the ratio of total cases to symptomatic cases. This would mean the CDC's worst case (which assumes 20% asymptomatic) is 0.8%, which is (as stated) lower than 0.86%. (their best estimate is 35% asymptomatic, but the "worst case" for death rate could come at the lowest asymptomatic ratio).

  • by hnjjz ( 696917 ) on Saturday May 30, 2020 @10:16AM (#60125156)

    That's about 0.26% of NYC's entire population of 8.4 million people. So even if all of NYC is infected, the infection fatality rate is still at least 0.26%. Antibody tests taken from NYC grocery store customers have ~20% positive rate. Assuming that's representative of the entire population, the actual NYC infection fatality rate is ~1.3%.

    • I'm not sure that's really NY's population - do they make sure to only count people *from* NYC proper and not greater metro area?

      Not that I think this bickering about exact percentages really matters.

      • by hnjjz ( 696917 )

        These are only the coronavirus deaths in NYC. NYC suburbs like Nassau and Westchester counties are counted separately and they also have high numbers of deaths.

        Also, the actual NYC population right now is likely quite a bit less than the 8.4 million estimate from before the pandemic. Large numbers of New Yorkers have fled the city and many apartment buildings in Manhattan are half empty.

    • That's about 0.26% of NYC's entire population of 8.4 million people. So even if all of NYC is infected, the infection fatality rate is still at least 0.26%. Antibody tests taken from NYC grocery store customers have ~20% positive rate. Assuming that's representative of the entire population, the actual NYC infection fatality rate is ~1.3%.

      And that ignores the large number of excess deaths in the area, at least some of which are directly caused by COVID-19 (as opposed to indirectly cause, e.g. by people avoiding going to the hospital when they need to, or suicides caused by depression from lockdown, etc.).

  • Cuomo'd (Score:3, Insightful)

    by Citizen of Earth ( 569446 ) on Saturday May 30, 2020 @10:21AM (#60125172)
    The death rate is probably pretty low if you don't catapult diseased corpses into nursing homes.
  • Hell hath no fury like activists who find out that the apocalypse they're promoting is not as apocalyptic as they first thought.

    • We're workin hard on getting a nation-wide riot going, does that help?
    • by HiThere ( 15173 )

      1) It's not an apocalypse if there is lasting immunity, or if a second case is milder than the first. (This is often true, though not always, but preliminary results indicate that it is true in this case.)
      2) It's still a major problem and needs dealing with properly.
      3) Because our medical systems are (at best) sized to deal with the normal level of problems, it requires strong measures to enable them to handle things without collapsing, yielding a much worse outcome.

      N.B.: The "lockdown" measures have prev

  • This article is really fucking stupid.
    No one is claiming that the *case* fatality is 0.26%. The CDC is saying that the *infection* fatality is 0.26.
    The *case* fatality in south korea is around 0.7. The infection fatality rate can only be lower.
    The number of *cases* is really of low informational value, because it is mainly determined be the number of tests and by who is tested.
    The 0.26 comes from antibody studies and matches the Age-Corrected result of the Heinzberg-Study among many others.

    • by valdyn ( 445073 )

      Addition:
      And the infection fatality rate deducted from antibody-studies does not take into account that some people do not develop the antibodies that are being detected (Maybe none, maybe different antibodies / cross immunity).
      I have seen estimates ranging from 30%-80% of people not developing those antibodies that are still immune.
      Primarly children are suspected not to develop antibodies.
      Overall fatality rate if everyone is probably as low as 0.05%

      • So, if New York State has 30,000 deaths. If the actual fatality rate is 0.26%, then the number of infected people in NY State is about 11.5 million, which is over 57% of the population. For the fatality rate to be 0.05%, it would mean that there would have to be 60 million infected in NY State. This is 3 times the population of the state.

  • The overall infection fatality rate is not very informative when outcomes depend so heavily on factors, such as age, access to treatment and general well-being. Different localities have different age distributions and risk factors. In New York City, people in the Bronx have fared much worse than those in Manhattan. In New York State, 20% of fatalities were nursing home residents, even though the nursing home population is nowhere close to 20%. It just doesn't make sense to use a single number (total death

    • Are you implying that amateur statisticians acting as armchair infectious disease analysts might not be the ones to listen to?
  • Use EITHER data question 1 or 2: (your choice)
    Data question 1: "Does what I'm hearing disagree with my personal feelings?
    Data question 2: "Does this information appear to validate Donald Trump in any way?"

    No: "It is a clear, concise, science-based and data-driven conclusion.
    Yes: "Clearly (source) has been politically compromised and is only spouting conclusions to serve a greater agenda.

  • So, the US numbers are a mess. Deaths are under-reported all over the place, the test numbers are conflated by CDC and many states. Misleading graphs by states using non-sequential dates.

    Then you have the president (the Head of State, the Big Cheese) denying there is a crisis for weeks, then advocates stupid cures, ...

    And Trump is blaming China?

    It looks to me US is trying the 1918 playbook - flu starts in US, don't tell anyone; wait for Spain to do something, call it the Spanish Flu. In the present
  • The Trump administration ruined the CDC.

    That is why The Lancet is calling for the CDC to be revived [thelancet.com].

Avoid strange women and temporary variables.

Working...