When 1 equals 2, the result is a retraction

bmcrnA group of psychiatric researchers in Norway has lost their 2013 paper in BMC Research Notes on the effects of antipsychotic medications on the brain after discovering that they’d botched their imaging analyses.

The article, “Does changing from a first generation antipsychotic (perphenazin) to a second generation antipsychotic (risperidone) alter brain activation and motor activity? A case report,” came from a trio of scientists at the University of Bergen and Haukeland University Hospital, also in Bergen. According to the abstract of the paper, which was published last May:

We present the case of a 53-year-old male with onset of severe mental illness in adolescence, ICD-10 diagnosed as schizophrenia of paranoid type, chronic form. We compared brain activation and motor activity in this patient during pharmacological treatment with a first-generation (perphenazin), and later switched to a second-generation (risperidone) antipsychotic drug. We used functional magnetic resonance imaging (fMRI) to measure brain activation and wrist worn actigraphy to measure motor activity. …

Our study showed that brain activation decreased in areas critical for cognitive functioning in this patient, when changing from a first to a second generation antipsychotic drug. However the mean motor activity level was unchanged, although risperidone reduced variability, particularly short-term variability from minute to minute. Compared to the results from previous studies, the present findings indicate that changing to a second-generation antipsychotic alters variability measures towards that seen in a control group, but with reduced brain activation, which was an unexpected finding.

Sometimes, however, the unexpected is unexpected for a reason. As the retraction notice states:

The authors have retracted this article as the fMRI data presented in the case report are incorrect. The activation data reported for session 1 are the activation data for session 2 and vice versa. As a result the discussion and conclusions of the case report are based on the wrong set of data and are no longer valid. The authors apologise for the error.

4 thoughts on “When 1 equals 2, the result is a retraction”

      1. I have detected several papers in which the authors have used the exact same figures, manipulated to indicate results of at least two different species. This is perfectly analogous to 1 = 2. How can two exact images be used to describe two separate species? Whether we are talking about microorganisms, animals or plants, such blatant fraud shouldn’t even be debated. It should be a straight retraction. Yet, despite formal complaints, the publisher remains in silence, as do the editors. In this case, what do RW bloggers suggest be done to take firmer action, not nly against the authors, but also against the passive, irresponsible nature of the editors and publsher? Can anyone with experience on the background of retractions indicate why it takes so long to deliberate when the evidence is so clear? What politics lie behind a decision to retract?

        1. I can answer this question. Two completely different trials, with different names, different designs (randomized versus accepter-versus-rejecter) and different sizes (289+289 versus 200+191)
          had completely identical results: same tables, same figures. Neither trial makes sense internally, by the way.

          The explanations for inaction given by journal editors are given in the supplemental Appendix 6,7,8,9:

          My personal favourite excuse is that notification of discrepancies must happen before a 3-week deadline. (statute of limitations?)

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.