Thoracic surgery journal retracts hypertension study marred by troubled data

The Annals of Thoracic Surgery has retracted a 2004 article by a group of Florida researchers who were found by their university to have misrepresented the provenance of their data.

If that construction sounds a trifle precious (er, weasel-y), that’s because the retraction notice does, too:

The Annals of Thoracic Surgery retracts the article …  following an investigation by the University of Florida, which uncovered instances of repetitious, tabulated data from previously published studies.

Tabulated we think we get: The paper,  “Effects of a new phosphodiesterase enzyme type V inhibitor (UK 343-664) versus milrinone in a porcine model of acute pulmonary hypertension,” does contain a lengthy table purporting to compare the new compound (an IV cousin of Viagra) with milrinone, an approved drug for heart failure sold as Primacor, in various measures of fluid and cardiac status. It’s been cited four times, according to Thomson Scientific’s Web of Knowledge.

But what’s the definition, exactly, of “repetitious data”?

The qualifier of “previously published studies” suggests plagiarism, as does the university inquiry. Again, however, without more information we’re left to assume.

We have reached out to the editor of the journal for answers to these questions, and we have tried to contact two of the paper’s authors. We’ll update this post when we learn more.

Meanwhile, a humble suggestion to journals: Fuzzy language in retraction notices doesn’t serve anyone’s interest — particularly not the readers. As we saw recently, a particularly vague editor’s letter can make people angry and damage the standing of the publication.

Please see an update with comments from the journal’s editor, and one of the retract study’s authors.

4 thoughts on “Thoracic surgery journal retracts hypertension study marred by troubled data”

  1. Thanks for highlighting ambiguous wording. It’s really important that retraction notices are clear, so they distinguish honest error from misconduct. Journal editors should work hard at this — although it’s not always their fault, as sometimes the institutional reports are written in such vague language it’s impossible to tell whether misconduct occurred or not.

  2. I can get the original article, but not the retraction: the latter requires a sign-in.

    From the wording, I’m guessing that the authors published the data, then reformatted it into a new table and published it again, and that’s a no-no. You’re not allowed to publish the same data twice.

  3. Definition of “repetitious data” –
    http://ats.ctsnetjournals.org/cgi/content/abstract/77/1/238

    The author used data from a “previously published study” for the article in question (both published in 2004).

    The author failed to disclose this. There also seem to be problems with the design of the study, but it’s hard to tell because the article is so short.

    In addition to being short, the article was terribly inconsequential. As someone who works in cardiac surgery, has read both articles in question, it was terribly inconsequential.

    Frightfully inconsequential. No disrespect to the author.

    I hope the author has learned how to better design studies.

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.