Yesterday I had a request to do some sleuthing on the article,”Androgens versus placebo or no treatment for idiopathic oligo/asthenospermia. Vandekerckhove P, Lilford R, Vail A, Hughes E. Cochrane Database Syst Rev. 2007 Jul 18;(4):CD000150. In the PubMed citation there is a nice big WITHDRAWN in front of the title. The doctor wanted to know why the article was withdrawn.
My first stop was The Cochrane Library on Wiley. According to The Cochrane Library,”This review has been withdrawn from The Cochrane Library as it has not been updated since 1996.” Ok makes sense, if the review article hasn’t been updated in that long then I can see withdrawing it. However, I began to look a little more and of course got a little more confused. Apparently another review article (same title, authors, and CD#) was published in 2000 and does not have giant WITHDRAWN printed in front of the title on the PubMed citation.
So my brain started to ask the questions…
- If the 2007 wasn’t updated since 1996, was the 2000 article updated?
- Why is it when I search for the 2000 article in PubMed there is no mention of it being withdrawn, but when I search The Cochrane Library for both the 2000 and 2007 review articles (both have the same CD#), the databse tells me it is withdrawn?
- Shouldn’t PubMed have a big ol’ withdrawn next to the 2000 citation too?
Another question that is bouncing around in my head is in the wake of so many scandals regarding scholarly publishing, were the 2000 and 2007 articles ever updated from 1996? The way The Cochrane Library has it listed it makes me think not. Because The Cochrane Library says the article hasn’t been updated since 1996 makes me believe that the original review article was written and published in 1996 or before. However when I search PubMed there are no articles by these authors on this topic before 2000. Yet when you look at the 2000 citation it clearly points to a 1996 article:
Cochrane Database Syst Rev. 2000;(2):CD000150. Review. Update in: Cochrane Database Syst Rev. 1996;(4):CD000150.
Obviously the 2007 review article has been withdrawn and I would be hesitant to use the 2000 one as well. But it is a little hard to figure out the story behind the withdrawal, other than it hadn’t been updated since 1996. But there are still a lot of questions left hanging out there. Now that things are going more digital it seems the breadcrumb path of the article is more nebulous which makes it difficult when those articles are no longer appropriate to use for treatment decisions.
I know the article hasn’t been retracted, it has been withdrawn because it is dated. However PubMed has noted that on the 2007 not the 200 citation which is extremely confusing. Neither a retracted article nor a withdrawn one are appropriate to use to treat patients. Therefore those situations should be noted more consistantly and effectively in the databases and there should be a clearer reason as to why something was retracted or withdrawn.
Without regard to my day job, this is not an official response from NLM.
This article has not been Retracted. Cochrane has withdrawn it from their database because it hasn’t been updated in whatever x year period their rules require. They are not saying the article was wrong at the time of publication. That’s what a retraction notice means. They are just saying this no longer meets their requirements.