The journal Neurology has issued an expression of concern for a paper linking shingles and stroke, which got press attention when it was published.
The journal’s note refers to “errors of data presentation,” which author Judith Breuer more narrowly defined as mistakes during transcription of a table. It’s unclear whether the results themselves – that herpes zoster, the virus that causes shingles, is a risk factor for stroke and other vascular problems – are being called into question.
Here’s the expression of concern for “Herpes zoster as a risk factor for stroke and TIA: A retrospective cohort study in the UK”:
With regard to the research article “Herpes zoster as a risk factor for stroke and TIA: A retrospective cohort study in the UK” by J. Breuer et al. (Neurology® 2014;82:206–212; published ahead of print January 2, 2014), we are publishing this Expression of Concern to alert readers that errors of data presentation have been uncovered since publication. The authors are preparing a corrected version of the paper.
Breuer filled us in:
A table was transcribed wrongly with the result that some values appeared in the wrong columns. Regrettably none of us picked this up before publication.
At least one other study has been published since that supports the findings of the the Neurology paper, which has been cited five times, according to Thomson Scientific’s Web of Knowledge.
We’ve asked Breuer when we can expect a correction, and if these issues affect the validity of the results. We will update with anything we find out.
Update 9:40 a.m. Eastern 12/17/14: Breuer got back to us with more information:
I am not sure when the corrected paper might be issued- the journal has not told me. No the table does not affect the validity of the results. All the discussions etc were based around the correct table and correct results. The problem was that the table was corrupted when it was pasted into the manuscript and this was not noticed by ourselves or by the editors. It was very unfortunate and a good lesson to us that we should scrutinise every part of a manuscript for typos/errors.