Hello, reviewers? How did a study manage to say it was randomized, but also that it wasn’t?

A journal has retracted a 2016 article for a litany of flaws, including plagiarism and a massive self-inflicted wound that should have obviated the first offense.  

According to the notice in Cardiology Research and Practice (a Hindawi title) for the paper, which compared two methods of threading a catheter into the heart’s arteries:

Cardiology Research and Practice has retracted the article titled “Coronary Angiography Safety between Transradial and Transfemoral Access” [1]. The article was found to contain a substantial amount of material from the following published articles:

-M. S. Kabir, M. M. Haq, S. R. Khan, M. Z. Chowdhury, M. L. Ali, and M. R. Karim, “Safety of radial vs. femoral artery access in coronary angiography,” Bangladesh Heart Journal, vol. 30, no. 2, pp. 68–73, 2016, doi: 10.3329/bhj.v30i2.28814 (not cited).

-Brueck, D. Bandorski, W. Kramer, M. Wieczorek, R. Höltgen, and H. Tillmanns, “A randomized comparison of transradial versus transfemoral approach for coronary angiography and angioplasty,” JACC: Cardiovascular Interventions, vol. 2, no. 11, pp. 1047–1054, 2009, doi: 10.1016/j.jcin.2009.07.016 (not cited).

That’s fairly ho-hum. Things get a bit more interesting in the next section:

There are also concerns with the reporting of the study design. The methods state “It was prospective, randomized, single-centre study conducted in the Department of Cardiology, LPS Institute of Cardiology, GSVM Medical College, Kanpur, U.P, India, where all cases of diagnostic coronary angiography (CAG) of 1997 consecutive patients for various reasons over a 12-month period (from January 2015 till the end of December 2015) were reviewed for this analysis,” but then that “The choice between transfemoral or transradial artery access was operator’s discretion with right radial approach being the default strategy.” These are mutually exclusive approaches to treatment allocation, that is, randomization versus surgeon discretion. Additionally, a trial registration number is not included.

Let that sink in for a second. A study described as “randomized” was not only not randomized at all, but was described as nonrandomized in the article. Although we can’t excuse the plagiarism — which of course is grounds for retraction — the journal might want to say a few mea culpas for missing an obvious and disqualifying discrepancy during peer review and the editing process.

The paper has not been cited since it was published, according to Clarivate Analytics’ Web of Science.

We asked Matt Hodgkinson, Hindawi’s ethics czar, for more information about the paper. He told us in an email that:

we were contacted in December 2016 about text overlap with one of the two articles mentioned in the notice. This was not major plagiarism, but when we re-ran similarity checks we found further overlap with another article. The overlaps are about 350 words with Kabir et al. in fragments throughout the text and about 70 words with Brueck et al. in fragments in the abstract.

Because of this I also reassessed the article (I am familiar with the critical appraisal of clinical trials) and found the reporting / study design issues that are included in the notice. The problem with patient assignment is obvious in hindsight, but I was checking the article because other issues had already been raised – and some of the text overlap is in the methods – and I can understand how such a contradiction can be overlooked. The peer review was handled by a cardiovascular surgery researcher with statistical expertise who used two cardiovascular surgery researchers as reviewers. The authors did not respond to our post-publication queries and the handling editor was unavailable, both of which caused delays. We consulted another editor, also a cardiovascular medicine researcher, who recommended retraction.

Hodgkinson added that Hindawi has asked the authors’ institution to get involved. And, he said,

We have revised our editorial screening since then to aim to ensure we require a registration number for all trials and we are planning to mandate some reporting guidelines…

For instance, Hodgkinson said the publisher plans to follow the guidelines established by Consolidated Standards of Reporting Trials (CONSORT), a 25-item checklist for randomized clinical trials.

Like Retraction Watch? You can make a tax-deductible contribution to support our growth, follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up for an email every time there’s a new post (look for the “follow” button at the lower right part of your screen), or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at team@retractionwatch.com.

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.