NHacker Next
  • new
  • past
  • show
  • ask
  • show
  • jobs
  • submit
Many scientists citing two scandalous Covid-19 papers ignore their retractions (sciencemag.org)
sndean 1195 days ago [-]
>> A lead author of the paper, biomathematician Maik Pietzner at the University of Cambridge, said that although the paper was submitted after the retractions occurred, it was written beforehand, and “the current pandemic requires immediate response.” However, the paper was published 4 months after its submission.

Particularly during the pandemic, the peer review / publishing process has become so slow in some cases that it's easily understandable to have not removed a citation, or to have had a year go by before publication.

1) I can't think of too many times that I've gone back to the source to read a paper. I download the PDF and never go back the website. How would I know about the retraction? 2) Changes to the text can require another round of review (possibly adds two months?). 3) Authors are generally against issuing corrections post-publication since the box that's added to the top of the paper looks similar to a retraction [0]. 4) Authors are busy or lazy.

>> Given that no editor or reviewer caught the problem, she said, “I plan to discuss with the staff incorporating such screening into manuscript processing.”

Shouldn't it be the responsibility of the publisher to add a notification to papers that cited a retracted paper? ("Be wary of things relating to ref #32.") We pay Nature Communications (cited in the article) ~$5000 to publish [1]... this should already be done for you.

[0] e.g., this PNAS paper: https://www.pnas.org/content/116/32/15877. Why not make the retraction notification a bright red box?

[1] https://www.nature.com/ncomms/about/article-processing-charg...

ta988 1195 days ago [-]
Zotero, a Free citation manager will tell you when a paper is retracted. it will also tell you (with a plugin this time) if it is discussed on pubpeer.
netizen-9748 1195 days ago [-]
that's a feature I was unaware of. I used EndNote for a while but I don't know if they also offer this service. Is Zotero the only one that does this? That's an insanely helpful feature for research
stareatgoats 1195 days ago [-]
I believe https://scite.ai/ has a similar service (might have to pay USD 15/month for it though)

Discussed on HN here https://news.ycombinator.com/item?id=23146091

mellosouls 1195 days ago [-]
This lack of timely response due to missing the retraction might be understandable in normal circumstances but it's difficult to see how any researchers in the area were unaware of these retractions as the scandal was major news outside of the field.
samizdis 1196 days ago [-]
Here is a list of retracted Covid-19 papers, as collated by Retraction Watch:

https://retractionwatch.com/retracted-coronavirus-covid-19-p...

danappelxx 1195 days ago [-]
andrewflnr 1195 days ago [-]
Wow, the comments under that are wild. Is is just me or is the fraction of crazy people in that thread pretty high?
toomuchtodo 1195 days ago [-]
It would be nice if retraction watch publicly messaged those who cite retracted papers and publicly logged their responses.
jansan 1195 days ago [-]
So why is not every paper that cites a retracted paper automatically flagged? THe peer reviewing process seems to have a lot of room for improvement.
matthewdgreen 1195 days ago [-]
Because the way papers are reviewed is highly manual and relies on individual reviewers who are acting as volunteers and don’t work full time doing review work. This makes it very difficult to implement systematic checks of the sort that would catch minor pieces of retracted work that aren’t central to this paper’s results. To fix this systematically you’d either need automated systems or human-executed “checkbox processes” with training.

Of course you could implement those systems, but they’d be costly either in terms of further centralizing the peer review process behind companies like Elsevier or taking critical human time away from reviewing the actual results. And the benefits if you implemented these review systems might be highly limited, plus you’d probably run into a host of additional problems.

On the “highly limited” side, many papers have a related work section that doesn’t list primary results that are relevant to the paper’s findings, but instead mainly serves as a reference guide. You might be writing a paper on (here I’m making up silly analogies) LED light bulbs, and you’d have a set of references that begins with “a line of work looks at incandescent light bulbs”. An accidental citation to a retracted work there is sloppy, but mostly scientifically irrelevant when no result in the paper actually refers to anything in those citations (which is why these citations slip by manual reviewers, who focus their valuable review time on citations that affect the scientific results in the paper.)

You also have the more annoying problem that citation formats aren’t standardized, so it’s easy for a retracted paper to be cited in a format that an automated review system will miss. Google Scholar throws all the might of Google’s systems at this, and still treats the preprint version and conference version of my papers as different works. (This might be the right treatment or it might be the wrong one, but it’s murder on retraction scanners if the author cites the preprint.) Plus then there’s the additional burden of dealing with unimportant retractions that are issued after a paper that cites them has been published, which should be addressed but largely just cause annoying work for scientists and publishers and (as the article notes) are aggravating because it’s difficult for readers to distinguish a minor correction regarding an unimportant reference from a major one that affects the findings of the work.

jansan 1195 days ago [-]
Science needs a standardized and open reviewing and publishing process. The current reviewing process may have been alright in the 80s, not so anymore.

Bill Gates, are you litening? This is a chance to revolutionize science!!!

chrisseaton 1195 days ago [-]
Citing a paper doesn't imply you think it's either good or bad. You have to read the text to understand what the authors say about it.
ta988 1195 days ago [-]
flagging doesnt mean forbidding, just a warning to make sure everybody is aware.
yawnxyz 1195 days ago [-]
editors and peer reviewers really need to do a better job flagging citations of retracted papers...
chrisseaton 1195 days ago [-]
Again, the issue is not the citation... it's what the citation is used to do. If an historian cites Mein Kampf it's not because they're agreeing with it, is it?
jjk166 1195 days ago [-]
If a historian cites Mein Kampf, they're using it as a primary source - they might not agree with the opinions expressed within but they do believe Hitler expressed those opinions.

If a historian cited a work that turned out to be a forgery, and did not acknowledge in their paper that it was a forgery, then their paper would certainly be suspect. True citing a retracted paper is not automatically 100% unacceptable, but it is something that should automatically demand further scrutiny.

chrisseaton 1195 days ago [-]
> If a historian cites Mein Kampf, they're using it as a primary source - they might not agree with the opinions expressed within but they do believe Hitler expressed those opinions.

Yes... same as a scientist uses citations.

Citing someone doesn't mean 'I fully believe and support this person', it means 'I'm referring to what this person said' and they believe that's what the paper said. That's all. The text using the citation adds the context of what the author thinks.

jjk166 1194 days ago [-]
You've completely missed what I was saying. A historian doesn't cite Mein Kampf to give context to what they're saying, the historian uses Mein Kampf as data. For a historian, Mein Kampf is perfectly valid data. On the other hand, Mein Kampf 2: Meiner Kampfer is not valid data, and any paper citing it as if it were should be immediately considered suspect. A scientist citing a retracted paper is like a historian citing a forgery: fine if it's a harmless aside, catastrophic if their claims hinge upon its reliability.
chrisseaton 1194 days ago [-]
> the historian uses Mein Kampf as data

So... the same way that a scientist could cite a retracted paper. To say that these people said this thing and then it was retracted. To discuss the context of that retraction and what it means, and so on.

jansan 1195 days ago [-]
Mein Kapmpf is not a scientific paper that can be retracted. Or have I messed some recent developments?
chrisseaton 1195 days ago [-]
> Mein Kapmpf is not a scientific paper that can be retracted.

I didn't say it was.

> Or have I messed some recent developments?

Why are you being sarcastic?

jansan 1195 days ago [-]
>Why are you being sarcastic?

It's my nature.

1195 days ago [-]
de6u99er 1195 days ago [-]
A couple of years ago I had an idea how to solve this issue.

IMO blocchain technology could be actually a solution for such a problem, where one publications cryptographic hash consists of the content of the publication and the cryptographic hashes of it's citations. If a citation get's invalidated for whatever reason, papers citing it would automatically get invalidated too, and so on. Until someone fixes or removes the faulty citation and revalidates their part of the chain, and so forth.

Just imagine authors having puplic wallets proving their authorship.

If someone wants to pick up this idea, I'd be happy with 10% of the earnings that can be made with it.

runsWphotons 1194 days ago [-]
I guess everyone is down on blockchain ideas but this seems cool.
rurban 1194 days ago [-]
Easy to explain: politics.

Amongst scientists everybody ignored those political papers, just as the many political models with doomsday predictions (also Lancet). They understand the need for politicians to be inaccurate and unscientific. Within the covid-19 community most flaws in preprint papers were immediately pointed out, but you could not blog about without loosing your job (Trump), but nevertheless news magazines still published the worst nonsense.

AntiImperialist 1195 days ago [-]
Anyone who believes vaccinations don't potentially cause autoimmune disorders, which includes some autism spectrum disorders, is a retard. If you know how vaccinations work, why the f would vaccinations not cause autoimmune disorders in a significant number of people?

Pharmaceuticals use their financial powers to constantly go after researchers who pursue studies that show these links.

mhkool 1195 days ago [-]
The biggest scandal so far: a retraction paper for the SARS-CoV-19 PCR test was submitted in November 2020 but ignored so far. The retraction paper is here: https://cormandrostenreview.com/report/
FabHK 1195 days ago [-]
Nonsense. So, here's what happened:

Drosten, Corman et al. from the Charité (among Europe's largest and most reputable university hospitals and medical schools) published a paper [1] in January 2020 [2] in the "Eurosurveillance Journal" in which they described a diagnostic workflow to detect SARS-CoV-2 (not "SARS-CoV-19", as you say) which they had developed and tested (both sensitivity and specificity). That became known as the Drosten PCR, and was the standard procedure initially to detect the virus (at least in Germany).

A group of 22 nutcases with bad English claim that the Drosten/Corman paper is severely flawed, put up a "report" "refuting" it (on a website, not published) and demand that Eurosurveillance retract it.

That report (which highlights a few very minor actual issues, but is otherwise false, misleading, and blown up entirely out of proportion) is later used by covidiots to claim that PCR testing is flawed and full of false positives, the virus doesn't exist, and further nonsense.

Needless to say, there are by now several different PCR testing protocols, they have been developed further, crosschecked, etc, and there is no major problem with PCR testing. Certainly no "big scandal".

This is political posturing and fabrication applied to medicine. Sad.

[1] https://www.eurosurveillance.org/content/10.2807/1560-7917.E...

https://pubmed.ncbi.nlm.nih.gov/31992387/

[2] It was submitted January 21st 2020 and accepted for publication on January 22nd 2020. On January 23rd 2020 the paper was online.

tboyd47 1195 days ago [-]
This comment is a bonanza of appeals to authority, straw man arguments, and ad hominems. At HN we generally prefer to see steelmanning. Would you like to take a shot at explaining why the most charitable interpretation of the Drosten review is misguided?
FabHK 1195 days ago [-]
It would be ad hominem if I said that they are nutcases with bad English, and therefore their arguments were invalid. It would be an appeal to authority if I said that the Drosten has received Germany's Federal Cross of Merit (twice), is co-discoverer of SARS-CoV (the previous one), has researched MERS-CoV, and has been called "one of the world’s foremost experts on coronaviruses" by Science magazine, and therefore the Drosten PCR is good.

However, I am saying that these guys are nutcases with bad English, and Drosten is an eminent expert on coronaviruses, and also the Drosten PCR has proven to be fine, and also the "report" the nutcases wrote has been largely refuted, by people that know much more about this than I do.

This article outlining the "Ad hominem fallacy fallacy" might be instructive.

https://laurencetennant.com/bonds/adhominem.html

> Put briefly, ad hominem is "You are an ignorant person, therefore your arguments are wrong", and not "Your arguments are wrong, therefore you are an ignorant person." The latter statement may be fallacious, but it's not an ad hominem fallacy.

tboyd47 1195 days ago [-]
You just repeated the same thoughts in an even more condescending way and still didn’t explain why they’re wrong.

It’s so odd that top level comment would be flagged.

rualca 1195 days ago [-]
On the contrary. The reply is a clear explanation of why the ad hominem accusation makes no sense.

The initial comment was rightfully flagged because it either ignores the bad and unscientific nature of the attacks on the paper.

1195 days ago [-]
thu2111 1195 days ago [-]
No. The replies don't address a single scientific point whatsoever, instead distracting with irrelevancies like attacking the English of the authors (which is plenty good enough), claiming the PCR test is "fine" when it's obviously nowhere even close to fine for the purposes it's being used for, and writing off things like the evident lack of review as a minor unimportant thing.

It's a classic example of desperately trying to shoot the messengers and tboyd47 is correct to say it doesn't belong here. His post was flagged simply because he's another messenger and thus is getting shot - it obviously has nothing to do with the politeness or quality of his posts.

rualca 1194 days ago [-]
> No. The replies don't address a single scientific point whatsoever

The reply actually addresses all the points.

* It provides an objective description of the test, supported by primary sources.

* Provides a description of the sources of criticism that were referred to in the initial statement.

* Describes the nature and quality of the criticism.

* Allows you to check that the papers under attack were not retracted, which is the testament to the scientific substance of said attacks.

* More importantly, raises the attention to the fact that there are far more test procedures nowadays which, even if we took the complains at face value, render them null.

There's a radical fringe which for some unexplained reason are both radically anti-science and heavily invested in trying to produce and use fraudulent pseudo-scientific work to attack scientific findings they find politically inconvenient, as a kind of Trojan horse based on appeals to authority.

thu2111 1194 days ago [-]
None of those points are convincing.

Firstly, lack of retraction does not imply quality. I've read papers about COVID that were not merely wrong but deliberatively deceptive, and they haven't been retracted. Public health research just has a huge problem with bad papers, and it's not one they're admitting to.

Secondly, have you actually read the criticisms? A big part of it is that the Drosten protocol appears to have been published via a journal that he himself controls, hence the 24 hour submission-to-publication turnaround time. Obviously if that is true the paper will never be retracted, thus using lack of retraction as a way to attack the critics is circular; the non-independence of the publication path is itself a part of their criticism.

Thirdly, I read the criticism. A few things were indeed quibbly and didn't seem that important, others were not good but perhaps understandable if you make allowances for extreme speed. Other criticisms were highly substantive. But this may depend on your pre-existing value system. A critical problem with PCR testing is the way the people behind it appear to consider false positives either non-existent or entirely free: it's clear that everything about these testing programmes is designed to maximize positive results at all costs. They're terrified of false negatives but don't care about FPs at all. That isn't excusable.

Finally, w.r.t. "more test procedures", they all seem to have very similar reliability problems that are dismissed with circular logic of the form, "COVID is anyone who tests positive, therefore there are no false positives" or "A PCR positive means it detected RNA, and PCR is the most sensitive way to detect RNA, therefore if RNA is present you get a PCR positive, therefore the test works fine", all of which is entirely blind to the way it's being used in the real world.

There's a radical fringe which for some unexplained reason are both radically anti-science and heavily invested in trying to produce and use fraudulent pseudo-scientific work to attack scientific findings they find politically inconvenient, as a kind of Trojan horse based on appeals to authority

Yes, that fringe is called the public health establishment.

tboyd47 1187 days ago [-]
Sad, but true. We're not anti-science. HN is one place where I'm quite sure everyone loves science. What we are is aware that the establishment entrusted with interpreting science has gone completely off the rails, and it has nothing to do with red vs. blue.
aaron695 1195 days ago [-]
Rubbish, it was the well thought out explanations I come here for.

You should have contributed something to the conversation other than saying steelman.

I picked a randomish name (English based since there will be more English articles) from the signatures and found this

Dr. Kevin P. Corbett

https://www.thelondoneconomic.com/news/who-were-the-so-calle...

eisstrom 1195 days ago [-]
I just checked the twitter account of the first author (who does not seem to be currently affiliated with any scientific institution). I can only understand his German and English tweets. He likes to push his own book, retweeted a post claiming "just stop testing for the virus and people will die of influenza again", and calls other peoples work pseudoscience.
mhkool 1195 days ago [-]
Why don't you talk about the 10 points that the scientists state to argue that the test is flawed? And what do you have to say about the other 21 scientists? What do you have to say about the former Pfizer chief scientist which is co-author?
morsch 1195 days ago [-]
Here's a fairly exhaustive rebuttal: https://mobile.twitter.com/BMauschen/status/1333466298072911...

I'll leave the translation up to the readers.

That former Pfizer guy apparently claimed in October that the coronavirus pandemic is “effectively over” in the United Kingdom. Clearly a man with special insights.

https://www.politifact.com/factchecks/2020/dec/02/blog-posti...

gus_massa 1195 days ago [-]
Autotranlation for the lazy: https://translate.google.com/translate?hl=&sl=auto&tl=en&u=h...

(For the lazier, the technical discussion starts at the 12th tweet.)

bigcorp-slave 1195 days ago [-]
This appears to be a totally unrelated set of authors attempting to force a journal to retract someone else’s paper. They literally have a dedicated website to promote this retraction and if you look up the authors, some of them headline their Twitter with it. The first author just today posted on Twitter that mRNA vaccines can alter your DNA.

The intent of the retraction seems to be to suggest that the epidemic is overblown, to which I would respond by pointing to the the two million dead people, and the excellent correlation of positive test results with new dead people.

briandear 1195 days ago [-]
Is the retracted paper correct? That’s the only question that matters. And perhaps those scientists have developed their views by actually looking at the science? Is that possible? Don’t shoot the messenger.
morsch 1195 days ago [-]
There is no retracted paper. Like the GP correctly points out, this is a website designed to pressure another group to retract their paper. The authors of said paper have not retracted it, although there were minor addendums (none of them in response to the website, I hasten to add).

The paper in question -- Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR[1][2] --, was the first paper (as far as I know, one of the first, certainly) outlining a protocol for a PCR test protocol for Sars-CoV-2, back in January 2020. In other words, these were among the first people who documented a way to detect whether a sample represented Sars-CoV-2. The authors have previously published about other coronaviruses like Sars-CoV-1 and MERS, e.g. [3] appears at first glance to be a well received paper. Here's [4] their paper on the detection of Sars-CoV-1 (or, back then, just Sars-CoV) via PCR.

[1] https://www.eurosurveillance.org/content/10.2807/1560-7917.E...

[2] https://scholar.google.com/scholar?cites=1445202628475605806...

[3] https://www.sciencedirect.com/science/article/pii/S006535271...

[4] https://www.eurosurveillance.org/content/10.2807/ese.17.39.2...

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact
Rendered at 04:57:05 GMT+0000 (Coordinated Universal Time) with Vercel.