A drop in deaths across all diseases is highly suspect. a part could be explained due to doctor visit restrictions caused by lockdowns and personal considerations yet death doesn't wait for a doctor's visit and would only come sooner. in 1-2 years time we'll have a proper graph of all deaths totalled and we'll know what the approximate impact covid had regardless of cause of death.
The mangled graph shown in the article was edited by the original authors, not the journalist.
Why is it suspect? When someone dies of influenza in a year, but has heart disease, do we say they died of heart disease or influenza? I suspect influenza is the cause of death. Statistically, over the past 100 years we haven't HAD a pandemic, but wow here we are with one and everyone is fast to say "ignore the covid numbers, the real data is heart disease because it's been that way forever." Well, forever just got pushed outta the way.
I understand that cause of death is multivariable, like the complicated reduction in life expectancy due to poor habits, city smog, individual effects of smoking, etc. But if a person has heart disease and influenza on the table, shouldn't they statistically be worse off if covid might be in the mix as well? Regardless of which is written on the death certificate, the amount of supposed deaths 'caused' by covid even by minimal estimates should be significantly larger than the margin of error on death records we have. Total deaths per year ( including absolutely everything) 'only' totals 3 million which means 200k covid deaths is already 7% of that.
the reported differences are interesting no matter how you cut it, either covid isn't as bad, or it is bad and people aren't dying of certain other causes. Alternatively the US really sucks at counting.
Did you read the article? The point of the study was that covid has had no affect on the overall number of deaths. This is in contrast to most publications which cite at least 250,000 deaths.
Did you read the article? The point of the study is to say that statistically this doesn't add up. Well, Covid is a once in every 100 years pandemic. Heart disease, influenza, and pneumonia occur every year. Trying to use statistical analysis is the wrong approach. To those who think this is your proof that covid was overblown, you haven't read the article.
In case you're serious: yes I have read the article.
Of course a one off non-peer reviewed piece isn't proof of anything but it does reference interesting data.
Heart disease, influenza, and pneumonia really are things that kill every year, so why don't you think they can be statistically analyzed? This year is different than previous years, but it seems odd if you're suggesting Covid is beneficial for heart disease.
Interesting, Johns Hopkins said they deleted the study because it "was being used to support false and dangerous inaccuracies about the impact of the pandemic", and not because it was incorrect.
Regardless of if it's correct, I think this has become such a political issue in the US that the American left would be very hard pressed to publicly accept this study's claims.
I really hope that statisticians are able to look past their own personal convictions and look at possible ways our assumptions may have been incorrect. (Personally speaking, I wear a mask, I don't think the coronavirus is a giant conspiracy by the lizard people etc. but I do recognize that the modern scientific community has a tendency to think one way, and then remain cemented in that way for a while.)
Why would the American Left need to? Look, it's obvious how easy this is to manipulate. Tell people that every year people die at higher rates of these four causes, but in 2020 we don't see that happen but the Covid numbers are where we saw the increase. So covid is really just no big deal and Trump was always right." Am I getting the gist of where you're going with this? That the American right was always correct? Laughable.
What this says is that in 2020 they didn't see the same numbers that they've always seen. Heart disease death counts aren't as high as they've always been. But wow, take a look at covid - those numbers are WAY up. To the observer who wants to say that covid isn't bad and is just the left's fever dream, this is the proof in the pudding.
To the intelligent, highly educated left, this says "Covid is the biggest cause of death in 2020. The right is just looking for something else to point at because Trump said so."
I believe the article is saying something, but it's not saying the right thing. Covid, in 2020, has been the biggest killer. Trying to identify an alternative cause is just wishful thinking.
It's become sadly common to suppress true research results because gatekeepers are afraid of the wrong people using them. It's gotten so bad that when I see an instance of egregious academic censorship, I think to myself "Yep. That's probably true.".
At both society and personal levels, the truth is always the best policy. Suppressing the truth, even for an ostensibly good reason, never ends well.
Maybe there's something to the censored article. Maybe there isn't. But whenever I see heavy handed censorship, I have to ask myself, "is there something here?"
There absolutely is a general issue about suppressing "true research results", as with government scientists who have been forbidden from speaking when that research opposes the interests of politicians (for one example, "Government Scientist Blocked from Talking About Climate and Wildfires" at https://www.scientificamerican.com/article/government-scient... ).
However, this doesn't appear to be true search result.
If it is indeed an erroneous conclusion published in a newsletter by JH students, and people implied ("Johns Hopkins published a study") that it was actual research result published by JH staff, then is there anything JH can do to suppress falsehoods that wouldn't be interpreted by the uninformed as heavy handed censorship?
We're talking about something published by a student in an undergraduate newspaper.
The information I've looked at is clearly wrong, as I've highlighted elsewhere in these comments.
I used "suppress" because you wrote "suppress true research results" - given your qualifiers, I assumed you thought suppression was appropriate for other contexts.
For example, isn't retracting a paper a way suppress the propagation of egregiously wrong research results?
"Briand was quoted in the article as saying, “All of this points to no evidence that COVID-19 created any excess deaths. Total death numbers are not above normal death numbers.” This claim is incorrect and does not take into account the spike in raw death count from all causes compared to previous years. According to the CDC, there have been almost 300,000 excess deaths due to COVID-19. Additionally, Briand presented data of total U.S. deaths in comparison to COVID-19-related deaths as a proportion percentage, which trivializes the repercussions of the pandemic. This evidence does not disprove the severity of COVID-19; an increase in excess deaths is not represented in these proportionalities because they are offered as percentages, not raw numbers.
Briand also claimed in her analysis that deaths due to heart diseases, respiratory diseases, influenza and pneumonia may be incorrectly categorized as COVID-19-related deaths. However, COVID-19 disproportionately affects those with preexisting conditions, so those with those underlying conditions are statistically more likely to be severely affected and die from the virus.
Because of these inaccuracies and our failure to provide additional information about the effects of COVID-19, The News-Letter decided to retract this article. It is our duty as a publication to combat the spread of misinformation and to enhance our fact-checking process. We apologize to our readers."
That's rather clear and unequivocal. Reading that, in conjunction with the original article and other science papers, it appears clear that none of that minimizes in any way the impact of the pandemic if you know how to read those sources properly. That is apparently not the case for everyone.
You cannot choose science when it suits you but faith when it does not. Atheists are the only people qualified to discuss the American left's receptivity to science, because the American right's dual belief system (faith plus science) disqualifies them, as they have said out loud that they believe in the supernatural. Therefore if you are a part of the right lamenting the lefts refusal to consider the study, you are disqualified, respectfully, as you are from all critique of all receptivity to all studies because you overtly believe in the supernatural. If you are a conservative atheist, I am all ears. I am absolutely receptive to the study being true, but that begs the question, how can both the study be true and the world scientific leaders have a consensus that there is not only net new death, but significant net new death due to covid? What would explain the discrepancy? A well run conspiracy? I ask not in mockery but I genuinely want to know what explains the difference?
[Meta] May I suggest editing the title to "Johns Hopkins published a study saying corona is not a big deal. They then deleted it."
I don't see "nbd" that often, and I suspect most Hacker News readers aren't going to realize what "nbd" is either. I opened the article because I thought "nbd" was some medical or scientific term.
This is apparently the result of an "assistant program director of the Applied Economics master’s degree program" downloading the CDC data and playing with it. It's not even a study, it's just someone putting together some graphs. It's not clear that they have any experience in interpreting this sort of data, which is probably important. Random economists poking around datasets they don't understand have a habit of coming to weird conclusions that disappear when looked at by subject matter experts.
I'm not sure it even counts as a study, where was it published? Is there a preprint somewhere? Is it just a PowerPoint presentation?
It's a student paper, the author of the piece appears to be a neuroscience undergrad... I'm sure they're a fine student journalist but it's not exactly peer review.
I don't know, there's no paper, it's just powerpoint slides.
I have previously poked around NYC-area death statistics and it looks like 1) a massive spike in deaths by any cause that looks like several 9/11s followed by 2) a regression to the usual rate of deaths once COVID was largely brought to heel in the summer months. I can't imagine that the total number of deaths in NYC this year could possibly be the same as in 2019.
But I'm not an expert.
Ed: glancing over the CDC statistics, when COVID-attributed deaths go up, all-cause mortality goes up, too... by approximately the same amount. If a large proportion of those COVID deaths were mislabeled heart disease deaths, why is that? I made these graphs with the same CDC data:
The charts "Weekly number of deaths (from all causes)" shows it's a big deal.
Or, select the dashboard "Excess deaths with and without COVID-19" and see that the number of deaths without COVID is pretty close to the expected number of deaths, and not a drop that would compensate for the COVID deaths.
You can also see that while the linked-to site claims "The CDC classified all deaths that are related to COVID-19 simply as COVID-19 deaths.", the CDC writes:
> For example, in some cases, COVID-19 may have contributed to the death, but the underlying cause of death was another cause, such as terminal cancer. For the majority of deaths where COVID-19 is reported on the death certificate (approximately 95%), COVID-19 is selected as the underlying cause of death.
Furthermore, the CDC write "deaths from all causes excluding COVID-19 were also estimated. Comparing these two sets of estimates — excess deaths with and without COVID-19 — can provide insight about how many excess deaths are identified as due to COVID-19, and how many excess deaths are reported as due to other causes of death."
Since the surface-level description contradicts easily available evidence, why do you believe the rest of the details are any more meaningful?
Your comments are merely ad hominem attacks, and have no place in rational discourse. There is no priesthood of "subject matter experts" who are the only ones capable of studying and interpreting mortality data from the CDC website. There is a very clear anomaly here, that all non-COVID causes of death have reduced at the same time and by the same amount as COVID deaths have increased in 2020 (as compared with all previous years in available history). This is indeed a very interesting observation that requires explanation, rather than suppression by "experts."
Ad hominem is simply a judgment made on credibility. One does the same thing when you look for a doctor. It’s not that an unofficial doctor has no merit, but rather that they have negative credibility.
Just how valuable is it to be a credible person, especially if you want to make an impact in the world, such as by being a doctor or researcher?
The headline was making an appeal to authority ("Johns Hopkins published a study..."). I was responding to that. Johns Hopkins didn't publish a study, a Johns Hopkins student newspaper published some slides.
It's not an ad hominem to try to deflate an appeal to authority by pointing out that the authority is not all it's cracked up to be, and that the extent of the supposed study is some charts.
Just to add to this, it was the general-purpose undergraduate newspaper (the News-Letter), posting an interview with the author of this "study". The News-Letter is not a scientific or economic journal, and the staff don't validate or endorse the academic accuracy of this type of content.
Additionally, the Applied Economics master’s degree program is an adult extension program located on a satellite campus. Its student admission requirements [1] are substantially less rigorous than a traditional master's program at one of JHU's main schools.
I agree the headline here is completely misleading, as it attempts to tie the content to other schools/divisions of JHU that have a much more solid academic reputation.
If that’s the case, why wouldn’t they invite a subject matter expert to debunk this analysis and clarify why the numbers seem to suggest there is no pandemic? Why take the video down?
Could the numbers be coming from a drop in influenza? Since we are socially distancing, influenza is probably way down seeing as how it’s less contagious than Covid.
The problem is that coronavirus got political. Not only did politicians get involved which means the probability of lies surrounding covid is 100% certain. That's what politicians do, they open their mouth and lies comes out. The problem is worse, the scientists and advisors also got political. Therefore people must question the veracity of the claims. That doesn't bring us to the current problem, the current problem goes much further.
The censorship over all discussion about covid unless it specifically goes along with the politician's story. Then a huge number of tyrannical measures 'for our safety' and it doesn't look quite right. It makes everyone question what's going on.
Want to fix the entire situation? Take the politicians out of the equation.
> Since COVID-19 mainly affects the elderly, experts expected an increase in the percentage of deaths in older age groups. However, this increase is not seen from the CDC data. In fact, the percentages of deaths among all age groups remain relatively the same.
This directly contradicts the excess mortality data published by CDC itself. [1] From the article it's unclear to me what evidence led to the claims presented. So far the article seems very unconvincing. Seems like John Hopkins has good reasons to take it down.
Also bear in mind that the excess mortality we've seen so far is with all the measures to limit the impact of the pandemic. It's fair to assume that excess mortality would be a lot higher without those. Our World In Data has a well explained international overview and analysis. [2][3]
A quick look at the website where this was published should discourage anyone from sharing links from it. I know the HN crew will downvote this because 'murica but whatever.
Wasn't that the graphing-around that said that people are dying from Corona but the total number of deaths isn't increasing correspondingly, and quite possibly not at all?
I found it quite interesting, but it sounded more like a reason to investigate than a result of investigation. For example, some non-Corona treatments are happening in much lower numbers than usually; how "should" that affect the number of deaths from the relevant causes and what numbers are actually observed? It's not sensible to draw conclusions about Corona from the number of other-cause deaths without also looking at the changes to other-cause reasons and treatments.
43 comments
[ 5.3 ms ] story [ 107 ms ] threadThe mangled graph shown in the article was edited by the original authors, not the journalist.
the reported differences are interesting no matter how you cut it, either covid isn't as bad, or it is bad and people aren't dying of certain other causes. Alternatively the US really sucks at counting.
If we didn't take proactive measures, deaths from Covid 19 would be a huge deal.
Of course a one off non-peer reviewed piece isn't proof of anything but it does reference interesting data.
Heart disease, influenza, and pneumonia really are things that kill every year, so why don't you think they can be statistically analyzed? This year is different than previous years, but it seems odd if you're suggesting Covid is beneficial for heart disease.
Clearly a counterfactual that you can easily say without the burden of offering evidence.
Regardless of if it's correct, I think this has become such a political issue in the US that the American left would be very hard pressed to publicly accept this study's claims.
I really hope that statisticians are able to look past their own personal convictions and look at possible ways our assumptions may have been incorrect. (Personally speaking, I wear a mask, I don't think the coronavirus is a giant conspiracy by the lizard people etc. but I do recognize that the modern scientific community has a tendency to think one way, and then remain cemented in that way for a while.)
What this says is that in 2020 they didn't see the same numbers that they've always seen. Heart disease death counts aren't as high as they've always been. But wow, take a look at covid - those numbers are WAY up. To the observer who wants to say that covid isn't bad and is just the left's fever dream, this is the proof in the pudding.
To the intelligent, highly educated left, this says "Covid is the biggest cause of death in 2020. The right is just looking for something else to point at because Trump said so."
I believe the article is saying something, but it's not saying the right thing. Covid, in 2020, has been the biggest killer. Trying to identify an alternative cause is just wishful thinking.
At both society and personal levels, the truth is always the best policy. Suppressing the truth, even for an ostensibly good reason, never ends well.
Maybe there's something to the censored article. Maybe there isn't. But whenever I see heavy handed censorship, I have to ask myself, "is there something here?"
However, this doesn't appear to be true search result.
If it is indeed an erroneous conclusion published in a newsletter by JH students, and people implied ("Johns Hopkins published a study") that it was actual research result published by JH staff, then is there anything JH can do to suppress falsehoods that wouldn't be interpreted by the uninformed as heavy handed censorship?
We're talking about something published by a student in an undergraduate newspaper.
The information I've looked at is clearly wrong, as I've highlighted elsewhere in these comments.
I used "suppress" because you wrote "suppress true research results" - given your qualifiers, I assumed you thought suppression was appropriate for other contexts.
For example, isn't retracting a paper a way suppress the propagation of egregiously wrong research results?
"Briand was quoted in the article as saying, “All of this points to no evidence that COVID-19 created any excess deaths. Total death numbers are not above normal death numbers.” This claim is incorrect and does not take into account the spike in raw death count from all causes compared to previous years. According to the CDC, there have been almost 300,000 excess deaths due to COVID-19. Additionally, Briand presented data of total U.S. deaths in comparison to COVID-19-related deaths as a proportion percentage, which trivializes the repercussions of the pandemic. This evidence does not disprove the severity of COVID-19; an increase in excess deaths is not represented in these proportionalities because they are offered as percentages, not raw numbers.
Briand also claimed in her analysis that deaths due to heart diseases, respiratory diseases, influenza and pneumonia may be incorrectly categorized as COVID-19-related deaths. However, COVID-19 disproportionately affects those with preexisting conditions, so those with those underlying conditions are statistically more likely to be severely affected and die from the virus.
Because of these inaccuracies and our failure to provide additional information about the effects of COVID-19, The News-Letter decided to retract this article. It is our duty as a publication to combat the spread of misinformation and to enhance our fact-checking process. We apologize to our readers."
That's rather clear and unequivocal. Reading that, in conjunction with the original article and other science papers, it appears clear that none of that minimizes in any way the impact of the pandemic if you know how to read those sources properly. That is apparently not the case for everyone.
I don't see "nbd" that often, and I suspect most Hacker News readers aren't going to realize what "nbd" is either. I opened the article because I thought "nbd" was some medical or scientific term.
I'm not sure it even counts as a study, where was it published? Is there a preprint somewhere? Is it just a PowerPoint presentation?
That is what any serious institution would do.
I have previously poked around NYC-area death statistics and it looks like 1) a massive spike in deaths by any cause that looks like several 9/11s followed by 2) a regression to the usual rate of deaths once COVID was largely brought to heel in the summer months. I can't imagine that the total number of deaths in NYC this year could possibly be the same as in 2019.
But I'm not an expert.
Ed: glancing over the CDC statistics, when COVID-attributed deaths go up, all-cause mortality goes up, too... by approximately the same amount. If a large proportion of those COVID deaths were mislabeled heart disease deaths, why is that? I made these graphs with the same CDC data:
https://imgur.com/a/YejafCE
And of course, there's the CDC's own dashboard on excess deaths, which lets you slice this more finely.
https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htm
The charts "Weekly number of deaths (from all causes)" shows it's a big deal.
Or, select the dashboard "Excess deaths with and without COVID-19" and see that the number of deaths without COVID is pretty close to the expected number of deaths, and not a drop that would compensate for the COVID deaths.
You can also see that while the linked-to site claims "The CDC classified all deaths that are related to COVID-19 simply as COVID-19 deaths.", the CDC writes:
> For example, in some cases, COVID-19 may have contributed to the death, but the underlying cause of death was another cause, such as terminal cancer. For the majority of deaths where COVID-19 is reported on the death certificate (approximately 95%), COVID-19 is selected as the underlying cause of death.
Furthermore, the CDC write "deaths from all causes excluding COVID-19 were also estimated. Comparing these two sets of estimates — excess deaths with and without COVID-19 — can provide insight about how many excess deaths are identified as due to COVID-19, and how many excess deaths are reported as due to other causes of death."
Since the surface-level description contradicts easily available evidence, why do you believe the rest of the details are any more meaningful?
Just how valuable is it to be a credible person, especially if you want to make an impact in the world, such as by being a doctor or researcher?
It's not an ad hominem to try to deflate an appeal to authority by pointing out that the authority is not all it's cracked up to be, and that the extent of the supposed study is some charts.
Additionally, the Applied Economics master’s degree program is an adult extension program located on a satellite campus. Its student admission requirements [1] are substantially less rigorous than a traditional master's program at one of JHU's main schools.
I agree the headline here is completely misleading, as it attempts to tie the content to other schools/divisions of JHU that have a much more solid academic reputation.
[1] https://advanced.jhu.edu/academics/graduate/ms-applied-econo...
https://www.youtube.com/watch?v=3TKJN61aflI
The censorship over all discussion about covid unless it specifically goes along with the politician's story. Then a huge number of tyrannical measures 'for our safety' and it doesn't look quite right. It makes everyone question what's going on.
Want to fix the entire situation? Take the politicians out of the equation.
https://www.euromomo.eu/graphs-and-maps
Is there a similar graph for the US? I couldn't find it on CDC website
This directly contradicts the excess mortality data published by CDC itself. [1] From the article it's unclear to me what evidence led to the claims presented. So far the article seems very unconvincing. Seems like John Hopkins has good reasons to take it down.
Also bear in mind that the excess mortality we've seen so far is with all the measures to limit the impact of the pandemic. It's fair to assume that excess mortality would be a lot higher without those. Our World In Data has a well explained international overview and analysis. [2][3]
[1] https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htm
[2] https://ourworldindata.org/excess-mortality-covid
[3] https://ourworldindata.org/covid-excess-mortality
I found it quite interesting, but it sounded more like a reason to investigate than a result of investigation. For example, some non-Corona treatments are happening in much lower numbers than usually; how "should" that affect the number of deaths from the relevant causes and what numbers are actually observed? It's not sensible to draw conclusions about Corona from the number of other-cause deaths without also looking at the changes to other-cause reasons and treatments.