We don’t usually cover “pretractions” (see #5 for why), but our friend Larry Husten over at Forbes has a story today about what appears to be a dead paper walking.
The article, in Heart, comes from a group of prominent researchers in Italy who have been arrested for possibly failing to adequately consent their patients, among other potential misdeeds.
According to Husten, the 2010 article in question, “A randomised trial of target-vessel versus multi-vessel revascularisation in ST-elevation myocardial infarction: major adverse cardiac events during long-term follow-up,” by Maria Grazia Modena (a past president of the Italian Society of Cardiology) and colleagues, may have been grossly misrepresented to the journal.
As Husten reports, a reader alerted him to potential problems with the article [several readers also submitted letters to the journal questioning the study’s conclusions]:
The questions raised about the paper lead to the inevitable suspicion that the trial very likely was not randomized and that patient consent may not have been properly obtained. …
The first and most obvious sign that something is seriously wrong with the trial is an extremely unusual imbalance in the number of patients in each treatment group. Normally in a randomized trial the number of patients in each group are roughly the same. (Proper randomization means that there will nearly always be small differences in the numbers.) But in this trial the sizes of the three groups were 84, 65, and 65. The paper in Heart contains no explanation for this imbalance. I have asked several clinical trial experts about this issue and they are in complete agreement that this imbalance is inexplicable and highly troubling.
Another red flag is an apparently perfect success rate in obtaining patient consent. A careful reading of the paper leads to the conclusion that all 214 consecutive STEMI patients with multivessel disease consented to participate in the trial. Again, this is unprecedented.
Perhaps those things can be chalked up to chance. But as Husten reports:
The article states that 263 patients were initially studied: 47 did not meet the inclusion criteria. But 263 minus 47 does not equal the 214 patients reported in the paper. Again, the article offers no explanation for this discrepancy.
The experts I spoke with wondered how this paper made it through the peer review process without resolving these questions.
Indeed. Or why the article was named an “Editor’s choice.” Or cited 48 times, according to Thomson Scientific’s Web of Knowledge.
Lindsey Fountain, journal manager for Heart, tells us:
We are currently investigating this paper and will take appropriate action as necessary once we have looked into it further.
So, let me get this straight. A group of Italian researchers have been arrested for perhaps fibbing about the consent rate of their patients. But, Dr. John Bohannon submits 304 (possibly 315) fake papers with fake identities, fake institutes, fake data sets and fake conclusions, causing undoubtedly massive psychological and possibly even financial injury to many journals and publishers, and his fraud is lauded by a publication in Science? Where’s the justice? According to my calculation, at least in Italian law, that should be 304 prison sentences to me.