A publisher makes an error in a publication about errors

Jennifer Byrne

Publishing a research paper is usually cause for celebration, after what is typically years of effort. Our recent paper in which we found that unexpectedly high proportions of papers in two journals described at least one wrongly identified reagent should have been no exception.

But alas. Any of our celebrations have been tempered by Springer Nature’s bizarre introduction of an unrelated figure into the paper. Here’s what has happened so far.

We initially posted the manuscript on bioRxiv in February 2023, and then submitted it to a journal. Unusually, we withdrew the manuscript after 4 rounds of revisions, as the editors required changes to our discussion of research paper mills that we did not accept. We then submitted the revised version to Naunyn-Schmiedeberg’s Archives of Pharmacology. The manuscript was accepted after further revisions and published online on January 9, 2024. 

However, on checking the published version, I was dismayed to find that Figure 1 had been replaced by an unrelated figure that we neither produced nor submitted for publication. The error was easy to spot: The correct Figure 1 compared the journal impact factors of Molecular Cancer and Oncogene from 2014-2021, while the substituted figure shows an unrelated Kaplan Meier plot. The proof returned to Springer Nature included the correct versions of all of the figures, so neither the authors nor the journal were responsible for the error.

I immediately wrote to Springer Nature about the error and copied the NSAP Editor in Chief, Roland Seifert. As I found the error on the day that the article was published, we hoped that the figure could be simply and quickly replaced. However, following email exchanges with Springer Nature representatives, it appears that: 

The problem here is that at the moment of online publication the paper becomes part of the scientific literature and the metadata of the publication are distributed to all relevant Abstracting+Indexing services. Though technically it would not be a problem to remove the paper from the website, we cannot do this since it is already officially published.” 

To date, while the publisher has now fixed the PDF, the incorrect figure remains visible in the freely available HTML version to users who have not cleared their caches. The publication has been accessed more than 300 times. Springer Nature has yet to provide an explanation of how the figure substitution occurred.

Clearly, a nonsensical figure in a paper about incorrect reagents is particularly unfortunate. It is similarly unfortunate that a publisher error that probably required minutes to achieve has no similarly quick path to correction. In the meantime, the authors’, the journal’s and the editor’s reputations risk being compromised, and time is wasted on communications that could have been devoted to celebrating this paper, or writing the next one.

While we are probably not the only group to have experienced a figure substitution by a publisher, some such errors may have gone unnoticed, as many authors might not expect or check for publisher errors. 

The possibility of figure replacements by publishers highlights another advantage of preprints- the correct version of Figure 1 remains online at bioRxiv.

Jennifer Byrne is conjoint Professor of Molecular Oncology and leads the PRIMeR group at the University of Sydney, Australia.

Like Retraction Watch? You can make a tax-deductible contribution to support our work, subscribe to our free daily digest or paid weekly update, follow us on Twitter, like us on Facebook, or add us to your RSS reader. If you find a retraction that’s not in The Retraction Watch Database, you can let us know here. For comments or feedback, email us at team@retractionwatch.com.

7 thoughts on “A publisher makes an error in a publication about errors”

  1. Dr. Byrne: I already suggested this on Twitter, but I think this is a situation where the APC should be returned to the authors/institutions who paid it, and Springer/Nature should offer an unreserved public apology to the authors and readers. The production team can take their own sweet time figuring out which process failed, or who pushed the button too soon. [I understand that fixing the publication record is not as easy as making changes to the website for the paper, but that is a different issue, and even then, not an impossible one].

    It doesn’t really matter to the outside world what series of errors led to this ridiculous situation, but in most business-to-consumer relationships, when a business screws up something in this way (“Oh, I’m sorry ma’am, I though you wanted the walls ‘Dungeon red’ not ‘Dungeness red’) it fixes the problem, refunds some/all of the fee, and apologizes, or else… the consumer will put a bad review on Yelp!

    Here, thankfully, you were able to find a platform with which to share the story with others. Thank you for sharing, and thanks to RW for helping. Good luck.


    1. Interestingly, I did not intend to include an emoji in my post, I typed something completely different on my PC (and don’t know how to insert an emoji deliberately even if I wanted to)… perhaps Springer/Nature or Retraction Watch has a sense of humor and are looking over my shoulder?

    2. Thank-you for this comment, and for your support on Twitter. While we’re committed to publishing open-access where we can afford to, there are mechanisms to protect authors from paying article publication charges (APCs) in cases such as ours. For example, my University only pays invoices when I confirm that the goods/ services have been received to my satisfaction. We therefore await the full resolution of this issue before proceeding with any payment.

  2. Waiting now for OMICS to use this line, the next time they’re caught fabricating papers and signing them with the identities of actual scholars, in order to fill in space in their journal-shaped garbage scows.

    “Oh dear, how sad, but it’s part of the permanent academic record now, so we can’t revise or retract it.”

  3. Depending on the origin of the publisher and applicable law, publishing a version that the authors did not agree to under the name of the authors could constitute a violation of the respective copyright law. In Germany, that would be the case here. With a law suit as option for the authors, the publisher might be motivated to make corrections rather sooner than later.

  4. Dear authors,
    I’m appalled to read that a wrong figure was uploaded and no apology or retraction or correction forthcoming from the journal.
    I must share an experience I recently had when submitting a ” Letter to the editor” to BMC Psychology. The original article we reviewed was published in BMC Psychology and hence our “Letter”to its editor.
    At first there was a lot of back and forth with its technical editors about putting on statements about ethical approval for the ” letter” and other nondescript issues. Just so the ” letter” got reviewed, we tagged along and made all the declarations which are not actually applicable to a ” letter to the editor” .
    Finally after a lot of back and forth, we got a letter from the chief editor that they couldn’t publish our ” letter” since they only published ” invited commentaries” . Either we didn’t get them or they did t understand what our “letter to the editor ” was about.
    Thank you for sharing your experiences.
    Hope the journal responds soon with something that will benefit the wider reading community .

  5. Thank-you for writing about the confusion that you experienced submitting a “Letter to the Editor” to a BMC journal. The confusion that you described reminded me of experiences published by David Allison and his team in 2016: https://www.nature.com/articles/530027a

    I should clarify that the incorrect figure in our paper has been corrected and a publisher correction was published on 16 January 2024, where the publisher took full responsibility for the error: https://link.springer.com/article/10.1007/s00210-024-02953-8. However, at the time of writing (01 February 2024) I still see the incorrect figure in the html version using my preferred browser, despite clearing my computer cache on several occasions. Others have reported the same problem, which indicates that online traces of publisher errors can persist post-correction. I hope that readers of our article will note the published correction and access the corrected pdf: https://link.springer.com/article/10.1007/s00210-024-02953-8, and that no further confusion will arise.

    While we recognise the publisher’s efforts to correct this error, we are yet to receive a specific explanation for how a graph comparing the journal impact factor trajectories for 2 Springer Nature journals was replaced by a Kaplan Meier plot of unknown origin. A description of the workflow that led to an author- and journal-approved figure being replaced by an unrelated figure could help to prevent similar errors from occurring in future.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.