Journal retracts IVF paper over data concerns

jargThe Journal of Assisted Reproduction and Genetics has retracted a 2010 paper by a group of Iranian researchers after concluding that the authors may have misrepresented both the nature of the study and the originality of the work.

The article, “Can fresh embryo transfers be replaced by cryopreserved-thawed embryo transfers in assisted reproductive cycles? A randomized controlled trial,” came from Abbas Aflatoonian, Homa Oskouian, Shahnaz Ahmadi and Leila Oskouian at Yazd Research & Clinical Center for Infertility at Shahid Sadoughi University of Medical Science. According to the abstract:

We designed this study to evaluate the effect of performing FETs instead of fresh embryo transfers (ETs) on ongoing pregnancy rates in IVF/intracytoplasmic sperm injection (ICSI) cycles. Clinical data suggest that cryopreservation of all embryos by vitrification and transferring them subsequently may be an effective strategy to enhance outcomes in assisted reproduction technology (ART). It was our aim to determine whether we could apply this new strategy to clinical practice

The paper has been cited 18 times, according to Thomson Scientific’s Web of Knowledge. But as the retraction notice indicates,

This article has been retracted at the request of the Editor and the ASRM [American Society of Reproductive Medicine] Publications Committee, based on the results of an investigation which found serious methodological flaws in the study.

What were those flaws? We spoke with David Albertini, editor of JARG, who told us that the case has been a complicated one. Albertini, who has edited the publication since its founding about five years ago, said the first problem that arose with the paper was that the authors had failed to register their trial on However, they did so between the initial submission and the subsequent revision, a listing which is available here. (They also seem to have published an abstract in Human Reproduction with the same title as the paper. That citation is available at Thomson Scientific, but not on PubMed or in Human Reproduction.)

Albertini said the authors had addressed the reviewers’ critiques and that he agreed to publish the paper, which he did, online, about a week after acceptance. But in meantime, a member of the publications committee

had found a paper that the same group had published in a different journal about a year before … It had some overlap in data content. It began to look like they were trying to republish the data.

What’s more, a comparison of the trial registration information with that in the two published articles — we can’t find the first one — revealed numerous “inconsistencies” that were sufficient to trigger an investigation by the publications committee. One of those, he said, was that the second paper did not in fact seem to involve a true randomized controlled trial.

Then the conclusion was made – with my input – that we should issue a retraction.

Albertini said the journal was in touch with the authors on at least two occasions, but that the researchers’ responses

were not satisfactory to explain the inconsistencies.

The authors are still fighting the decision, he said.

I’m getting an email a week from the people who published this paper. They’re begging me to reverse the decision, which I’m not going to do.

0 thoughts on “Journal retracts IVF paper over data concerns”

  1. It is indeed a poor paper. Here are a few of the more obvious problems/inconsistencies between trial registration and publication.

    As noted above the trial was registered AFTER completion but this not mentioned in the paper. Planned sample size 500 on registration document. Reported sample size 374 (187 per group). Primary endpoint in the registration document “implantation rate at 4 weeks”. Primary endpoint in the paper “ongoing pregnancy rate at 12 weeks”.

    The implantation rate is reported in the paper as “100/404 (24.7%) and 72/416 (17.5%) in the FET and fresh ET groups, respectively (p < 0.05)" i.e. a completely different denominator in both groups!

    Finally there was no mention of randomisation in balanced blocks but they ended up with two exactly equal groups, every patient got their allocated treatment and there were no losses to follow up! That should have set alarm bells ringing.

    It's wonderful to see these glaringly problematic trials retracted. I wish it happened more. It's not confined to less developed country researchers. Here's an example of a trial run by NIH-funded researchers from Boston which (among many other problems) changed from thee arms to two between registration and publication!

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.