That must have been what happened to a quintet of authors from Shanghai who’ve just had to retract an article from the Journal of Antimicrobial Chemotherapy. Here’s the notice
(sadly, behind a paywall) [see note at end of post] for “Role of clofazimine in the treatment of multidrug-resistant tuberculosis: a retrospective observational cohort assessment:”
After online publication of this article, it was brought to our attention by the Journal that an overlapping article had been published by us almost simultaneously in another journal1 and that there was no cross citation between the two articles.
We regret any problems our article and actions may have caused and we retract it from the literature.
The “overlapping article,” called “Clofazimine in the treatment of multidrug-resistant tuberculosis,” was published in Clinical Microbiology and Infection, and has been cited twice, according to Thomson Scientific’s Web of Knowledge. Here’s its abstract:
Clofazimine has shown activity against Mycobacterium tuberculosis, including multidrug-resistant strains in vitro and in animal studies. However, clinical experience with clofazimine in multidrug-resistant tuberculosis (MDR-TB) is scarce. We reported our clinical experience with 39 MDR-TB patients treated with combination regimens that included clofazimine. From January 2008 to March 2011, 39 patients received clofazimine for the treatment of MDR-TB in Shanghai Pulmonary Hospital. Patients had isolates resistant to a median of six drugs (range, 2–11 drugs). Of the 39 cases, 36 had cavitary changes noted on initial chest radiograph or chest computed tomography, and positive sputum-smear microscopy results at the time of MDR-TB diagnosis. At data censure, 15 of the 39 patients had successful therapy, with at least five consistently negative cultures documented for the final 12 months of treatment. Eleven continued to receive treatment. There were no deaths. Thirteen patients had a poor outcome, including four defaults and nine treatment failures. Culture conversion occurred in 22 cases at a median of 12 weeks. Side-effects occurred in 34 patients, mainly including skin discolouration, ichthyosis and gastrointestinal adverse events. No patients reported significant toxicity likely to be attributable to clofazimine therapy. Adverse events were managed by combinations of dose adjustment and symptom management. In our experience, clofazimine was well tolerated and may have efficacy in the treatment of MDR-TB.
And here is the abstract of the now-retracted article:
Background Clofazimine is used to treat patients with multidrug-resistant tuberculosis (MDR-TB)/extensively drug-resistant tuberculosis (XDR-TB), although clinical data on its safety and efficacy are lacking.
Methods We performed a retrospective observational cohort study evaluating the role of clofazimine in MDR/XDR-TB treatment in the Shanghai Pulmonary Hospital. Efficacy evaluation compared endpoints of 30 clofazimine-treated cases against 100 non-clofazimine-treated cases.
Results Out of 144 MDR/XDR-TB patients, 44 were treated with clofazimine. Of these, none experienced major side effects attributed to clofazimine requiring discontinuations. However, 20 patients had the clofazimine reduced to manage adverse events. Outcomes were similar in patients treated with and without clofazimine, although clofazimine-treated cases had a more frequent history of previous treatment with a fluoroquinolone and an injectable agent (43.2% versus 20%, P < 0.05). Patients treated with clofazimine took a long time for sputum culture conversion and achieved a low percentage of conversion overall.
Conclusions Clofazimine did not improve the chance of bacterial conversion, providing a chance of treatment success in MDR-TB cases. However, 100 mg of clofazimine once daily added to an individualized multidrug regimen was well tolerated in our study. It might be used as an alternative drug for the treatment of complicated MDR/XDR-TB patients.
So: Significant overlap of trial subjects, but hardly identical papers.
In cases like these, its often useful to examine the publication history. So let’s. The CMI version was submitted on August 1, 2011, with a revision submitted on October 18, 2011. It was accepted on November 3 of that year, and published on November 7. It wasn’t until a month later — on December 8 — that the authors submitted the now-retracted JAC version.
So how was it that the authors had to have the journal bring the duplication to their attention? One potential clue is that there were only three authors on the earlier version, while there were five on the later one. But the papers have the same corresponding author, so we’ve asked him for details and will update with anything we learn.
In the meantime, we are, as is often the case, reminded of this famous scene from Casablanca:
Update, 4:45 p.m. Eastern, 1/24/14: Oxford tells us the paywall in front of the notice was an oversight, and that they’ve removed it.