Epilepsy researcher gets retraction, correction after former colleague flags work

EpilipsiaEpilepsy researcher Toni Schneider has received a retraction and a correction in quick succession, after a former colleague raised red flags about the work. 

The retraction for Schneider, based at the University of Köln in Germany, is for “unintentional inclusion of erroneous data” due to limitations of the recording system used in the paper, according to the notice. 

Marco Weiergräber, a former colleague of Schneider’s, has claimed that the authors of the paper did not use the test properly. The journal editor, however, told us he believes the original analysis is an “honest mistake,” and there is “no evidence” to suggest that the authors intentionally published incorrect analyses.

Here’s the retraction notice, issued by the journal in March 2016:

Cav2.3 (R-type) calcium channels are critical for mediating anticonvulsive and neuroprotective properties of lamotrigine in vivo’ by Dibué, M., Kamp, M. A., Alpdogan, S., Tevoufouet, E. E., Neiss, W. F., Hescheler, J. and Schneider, T.

The above article, published online on 17 June [1] in Wiley Online Library (wileyonlinelibrary.com), has been retracted by agreement between the journal Editors in Chief, Gary W. Mathern, Astrid Nehlig, and Michael Sperling, and Wiley Periodicals, Inc. The retraction is due to the unintentional inclusion of erroneous data due to the limitations of the recording system used. The system demonstrated greater than a 99% loss of amplitude of the signal, and a non-linear response of the recording system to changes in amplitude of an input signal. Because the signal reported in the published manuscript is within the range of noise in the system, and validity of the signal cannot be assured, the editors have concluded that the paper should be retracted.

The 2013 paper has not yet been indexed by Thomson Reuters Web of Science.

And here’s the correction notice that was issued — also in March 2016 — for a 2014 PLOS ONE paper, “Evaluation of a Murine Single-Blood-Injection SAH Model,” which has been so far been cited twice:

The article reports the recording of telemetric data from implanted TL11M2-F20-EET transmitters. These results are described in Figures 4 and 5.

The TL11M2-F20-EET transmitter has a reported transmission bandwidth of 1-50Hz and a sampling frequency of 250 Hz. The maximum identifiable signal for the device is 125 Hz.

The article reports experiments measuring signals of up to 500Hz. Given the limitations of the transmitter employed, the results reported for frequencies beyond 100Hz are not reliable. We are thus issuing this notification to alert readers that the results in Figure 4 and Figure 5C reporting signals over 100Hz are not reliable.

Weiergräber, who worked in Schneider’s lab until 2010 but is now based at Federal Institute for Drugs and Medical Devices (BfArM) in Bonn, Germany, told us earlier this month his concerns about the newly retracted paper:

I pointed to the use of a radiofrequency transmitter out of its technical range that leads to analysis of data that could not be recorded. I’m an expert in the field of EEG radiotelemetry and when reading their recent new publications I became aware the new coworkers in Prof. Schneider’s lab were obviously doing something wrong here.

After creating PubPeer posts for the three papers and not hearing from the paper’s authors, Weiergräber began sending letters to the journals. He clarified that he has no personal problem with Schneider or his group, but he saw it as his duty to point out errors in literature. 

In July 2015, the authors of the now-retracted paper published a response to Weiergräber in Epilepsia. It reads, in part:

In his letter titled “Cav2.3 R-type Ca2+ channels in anticonvulsive treatment,” Dr. Weiergräber challenges the accuracy of our electrocorticography (ECoG) recordings in the high-frequency range and use of F20-EET transmitters for this purpose. Dr. Weiergräber erroneously guesses that the second channel of the F20-EET transmitters was used to record ECG studies that are not shown in the publication. The second channel of the transmitter was not used at all as leads of this channel had become too short to implant. Dr. Weiergräber correctly states that the actual bandwidth of the transmitter is 1–50 Hz and that accuracy in this range is greatest. Indeed, beyond 50 Hz, readable amplitudes decrease resulting in a decreased signal-tonoise ratio. However, because high-frequency oscillations are often read as low-amplitude currents, despite lower accuracy, some investigators including ourselves choose to include the nominal bandwidth (fivefold or greater) when using these transmitters to gain insight into higher frequency bands and to sample at a higher rate to avoid aliasing.

Schneider has not responded to a request for comment.

However, Weiergräber told us the journal informed him in November 2015 that it was going to retract Schneider’s paper; it was eventually retracted in March 2016. Weiergräber, however, said the retraction notice issued for the paper was “scientifically nonsense,” adding:

The Editors insinuate that the telemetry system has drawbacks, but this is wrong. The authors simply did not use it correctly. I contacted Epilepsia again to tell them that their explanation protects the authors blaming the telemetry system now. But the EiC said that the issue is closed.

Michael Sperling, co-editor-in-chief of Epilepsia and a physician at the Thomson Jefferson University Hospital in Philadelphia, Pennsylvania, told us: 

Dr. Weiegraeber contacted us to raise concerns about the validity of the results reported in this paper. We investigated the issue that he raised, and offered the authors opportunity to rebut the criticism and demonstrate that the findings were valid. We consulted an outside expert to ensure a fair hearing. After review of the original paper, we came to the conclusion that the original report was in error and that the results could not be substantiated because of limitations in equipment capabilities — the equipment could not measure the signal that the authors thought they were measuring.

He added:

We allowed the authors of the study to conduct [a] new experiment (not a clinical experiment as alleged by Dr. W) to determine whether the original findings might be valid. The new experiment, which took time to conduct, confirmed our opinion that the papers reported electrophysiology results that could not be measured by the equipment employed by the investigators. Once the second experiment was completed, we retracted the paper.

Sperling said he believes the original analysis is an “honest mistake,” noting that there is “no evidence” to suggest that the authors intentionally published incorrect analyses:

The retraction was published as soon as we were certain that it was justified.

Sperling added:

Our explanation of the retraction was accurate and sufficient for its purpose. We make no insinuations nor do we protect the authors.

Gary Mathern, the other editor-in-chief of Epilepsia from the University of California, Los Angeles, said the authors of the paper were given the opportunity to voluntarily retract the paper or provide evidence to the contrary. The authors, said Mathern, insisted they were correct based on manufacture notes, and did not provide their data when requested.

The letter Weiergräber sent to Epilepsy Research regarding another of Schneider’s papers was withdrawn after being published. Leonid Schneider (no relation to Toni Schneider) has posted a link to the original letter on his website. Weiergräber told us:

One day after my letter was online as early view he [the EiC of Epilepsy Research] told me that he will withdraw it (it was not a retraction to be technically correct) because he thought I have a conflict of interest. He did not tell me which conflict of interest that was supposed to be.

Weiergräber added that he has “no competing interests” with the Schneider group, which he left on good terms:

I simply had the possibility to become group leader at the Federal Institute for Drugs and Medical Devices in Bonn and I took this position in May 2010. Since that time I had no cooperation or personal contact with the Schneider Group.

We’ve reached out to the editor of Epilepsy Research, David Treiman, who is based at the Barrow Neurological Institute in Phoenix, Arizona. We’ll update the post with anything else we learn.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.

4 thoughts on “Epilepsy researcher gets retraction, correction after former colleague flags work”

  1. The reason why I (and others) feel unhappy about the Epilepsia retraction notice is that it states: “The retraction is due to the unintentional inclusion of erroneous data due to the limitations of the recording system used.” This has raised some questions in the telemetry field whether the system itself has drawbacks. I repetitively pointed to the fact that every recording system has its technical specifications, i.e.transmitter bandwidth, nominal sampling rate and Nyquist limit that have to be considered and were ignored in the aformentioned publications (Dibue et al., 2013, Kamp et al., 2014, Dibue et al., 2014).
    Unfortunately, the retraction notice is vague. It remains unclear to the scientific audience where the information of “99% loss of amplitude of the signal” is actually originating from. Does this hold true for the total bandwidth or does this show up from a specific frequency (range)? Where is this information coming from? Who measured this? For telemetry users the explanation is not referenced, not based on company (DSI) data and thus highly irritating. The violation of the Nyquist limit is not addressed at all (in contrast to the PlosOne comment).
    Finally I was repetitively seeking for help from Epilepsia but also PlosOne and the authors in answering the question:
    How can one find alterations in treated animals (Dibue et al., 2013, Kamp et al., 2014) in a frequency range (125Hz -500Hz) that obviously cannot be reconstructed while the controls remain unaltered in this frequency range?

    There has been no answer.

  2. The editors and the journal retracted, not the authors? Is that right? Amazing…..
    Even the PLOS One correction comes from “The PLOS ONE Editors”, not the authors?

    Probably not everyone is aware of the Nyquist-Shannon theorem. But virtually everyone with a normal school education has heard of digital audio on a CD and how it’s sample rate of 44.1 kHz limits it’s frequency range to around 20 kHz. And isn’t it common sense that you cannot reliably reconstruct an analog signal by picking a sample once only every other wave (= “measuring” a 500Hz signal with 250 Hz sample rate)?

    And the authors go out of their way to explain that everything is fine? They did not retract themselves?

    And indeed, the question that Marco Weiergräber asks as to how likely it is to generate data differences between controls and subjects when “measuring” frequencies in the range of 125Hz-500Hz using a 250 Hz sample rate is intriguing.

  3. This is a follow-up to the upper report.
    Elsevier has now published an updated version of my letter to the Editor in Epilepsy Research (Epilepsy Res. 2016 Dec 8. pii: S0920-1211(16)30377-1. doi:
    10.1016/j.eplepsyres.2016.12.003. [Epub ahead of print]) in which I pointed to the violation of the Nyquist-Shannon limit in Dibue et al. (2014).
    There is a response from the senior author of Dibue et al. (2014) (DOI: http://dx.doi.org/10.1016/j.eplepsyres.2016.12.004).
    The withdrawal notice of my original letter from April 2015 is said to be removed by Elsevier.
    As the violation of the Nyquist-Shannon limit and the use of the transmitters out of their technical specifications (Dibue et al., 2013; Kamp et al., 2014, Dibue et al., 2014) have been confirmed independently several times now (Epilepsia, Plos One), it seems senseless to comment again on what is mentioned in the response letter.

  4. Wow…the retracted Epilepsia paper was republished (even with higher impact)…what a success for the authors and what a tragedy for science…didn’t Epilepsia state that the data are invalid! How to republish invalid data? What a shame! and what a bad day for science.

    Dibué-Adjei M, Kamp MA, Alpdogan S, Tevoufouet EE, Neiss WF, Hescheler J, Schneider T. Cav2.3 (R-Type) Calcium Channels are Critical for Mediating Anticonvulsive and Neuroprotective Properties of Lamotrigine In Vivo. Cell Physiol Biochem. 2017 Nov 24;44(3):935-947. doi: 10.1159/000485361.

    Dibué M, Kamp MA, Alpdogan S, Tevoufouet EE, Neiss WF, Hescheler J, Schneider T. Cav 2.3 (R-type) calcium channels are critical for mediating anticonvulsive and neuroprotective properties of lamotrigine in vivo. Epilepsia. 2013 Sep;54(9):1542-50. doi: 10.1111/epi.12250. Epub 2013 Jun 17. Erratum in: Epilepsia. 2016 Nov;57(11):1947. Retraction in: Epilepsia. 2016 Nov;57(11):1947.

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.