Journal retracts — and republishes — small study on gamma rays for OCD

Screen Shot 2015-10-28 at 8.40.00 PMJAMA Psychiatry has retracted and republished a paper on a cutting-edge procedure for patients with obsessive compulsive disorder.

In the original paper, the authors claimed that three out of eight patients who underwent a procedure that used gamma rays to kill brain cells showed improvements 12 months later (versus zero in the group who underwent a “sham” procedure). But after a reader noticed an “inadvertent” error in the calculation of how many patients had improved, the authors realized that only two of the patients had responded meaningfully to the procedure.

The new results “did not reach statistical significance,” the authors write in a “Notice of Retraction and Replacement.”  JAMA Psychiatry published it yesterday, along with a new version of the articlea letter from psychiatrist Christopher Baethge pointing out the error, and an editorial. The original article is available in the supplemental material of the new version, with the errors highlighted.

Here’s the note in full for “Gamma ventral capsulotomy for obsessive-compulsive disorder: a randomized clinical trial,” which explains the error:

On behalf of our coauthors, we are grateful to Dr Baethge for his detailed reading of our study. He is correct that for the primary outcome, we inadvertently misclassified a patient (ATa3) as a responder when indeed that patient was a nonresponder. When preparing the table, we inadvertently included in our computation of treatment response a Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) score of 36 (the patient’s first measure) instead of 30 (the second measure). We defined the baseline Y-BOCS score for this study as the last value obtained before surgery. At that point, all patients knew they would receive the procedure.

Thus, instead of 3 of 8 patients responding in the treatment group at 12 months, 2 of 8 responded compared with 0 of 8 in the control group. Because of this error, the analysis has been redone and we have requested that JAMA Psychiatry retract and replace the original article.

In the corrected article, corrections have been made to the Abstract; Results, Discussion, and Conclusions sections of the text; Tables 2 and 3; Figure 2; and eFigure 3 and eTables 5, 6, and 7 in the Supplement. The article now concludes, “In this preliminary trial, patients with intractable OCD [obsessive-compulsive disorder] who underwent GVC [gamma ventral capsulotomy] may have benefitted more than those who underwent sham surgery although the difference did not reach statistical significance. Additional research is necessary to determine if GVC is better than deep-brain stimulation.”

We regret the errors caused by this misclassification as well as the confusion it caused for JAMA Psychiatry, readers, and potentially patients.

The original paper has been cited six times.

The abstract notes that various forms of the procedure have been around for a while, but formal studies hadn’t confirmed they were effective:

Select cases of intractable obsessive-compulsive disorder (OCD) have undergone neurosurgical ablation for more than half a century. However, to our knowledge, there have been no randomized clinical trials of such procedures for the treatment of any psychiatric disorder.

The original abstract concluded that the treatment in question — which directs gamma rays to a part of the brain to kill the cells underlying the disorder —  could benefit some patients:

Gamma ventral capsulotomy benefitted patients with otherwise intractable OCD and thus appears to be an alternative to deep-brain stimulation in selected cases. Given the risks inherent in any psychiatric neurosurgery, such procedures should be conducted at specialized centers.

Medscape covered the original study: “Gamma Knife Brain Surgery Promising for Severe OCD.”

The letter from Baethge explains that he discovered the error because the authors were “transparent” and provided a “wealth of outcome data.” He also points out that it’s now unclear whether the procedure has any benefits for patients with OCD:

[C]hance cannot be ruled out as a factor in explaining the results.

The corrections do not lessen the merit of this exceptional study, yet the bottom line of the randomized clinical trial may be this: We cannot be sure, but it seems as if GVC reduced symptoms in treatment-resistant obsessive-compulsive disorder. Two of 8 patients responded to treatment, and another patient developed delirium, likely as a consequence of GVC. Depression, anxiety, and quality of life were not improved.

The editorial explains what happened after Baethge discovered the “inadvertent” error:

After an error was discovered by Baethge, as also reported in a letter herein, the authors reviewed the data and confirmed an important but inadvertent error had occurred.

The editorial also explains the decision to retract:

Retractions are typically reserved for articles that have resulted from scientific misconduct, such as fabrication, falsification, or plagiarism, or from pervasive error for which the results cannot be substantiated…Publication of pervasive incorrect data resulting in a major change in the direction or significance of the results, interpretations, and conclusions, as occurred with the trial reported by Lopes et al, is a serious matter. However, in this case, the error was inadvertent and the underlying science is still reliable and important. Thus, we now publish this notice of retraction and replacement with explanation from the authors and a corrected replacement article as we believe it is important for readers, investigators, and clinicians to have access to correct results of this trial. We have included a version of the original retracted article showing the original errors and a version of the replacement article showing what was corrected in the online supplement of the corrected replacement article.

Indeed, a significant portion of retractions result from honest error, not misconduct. For example, we recently saw another case where a paper was retracted due to an error: A short line of code felled a widely reported finding that the risk of divorce increases when women fall ill, but not when men do. The journal retracted the study, and then republished a corrected paper and an editor’s note several weeks later.

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, and sign up on our homepage for an email every time there’s a new post. Click here to review our Comments Policy.

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.