Reader Comments

Post a new comment on this article

Registerig reviews/meta-analyses is a good idea

Posted by tkindlon on 08 Feb 2010 at 23:04 GMT

This was an interesting piece.

I think the point about registering reviews/meta-analyses is a good one so that we don't just get to see ones that might suit somebody's agenda.

Often if a particular approach has become standard practice with a particular condition (for example), people can be making their living from offering that approach so there can be a reluctance or pressure not to publish results which "rock the boat"and similarly there can be an incentive to publish results which re-inforce the current situation.

When reading an article earlier this year that was published nearly a decade ago (not a PLoS journal), I noticed the authors referred to doing a follow-up study. So I contacted the PI to find out had the results ever been published. They told me that the follow-up study had been done but they did not publish them. This is a very frustrating thing to hear. (This person was very obliging with various queries I had so I don't particularly want to name them).

But in general, what can be done in situations like this?

One thing that helps is when journals (like Plos Medicine) allow online comments.

It is time-consuming writing a letter that is good enough for publication and one doesn't have a guarantee that it will be published by an editor. But with online comments where the standard of a submission does not have to be so high (not just in terms of the quality of the points but also because it is work to make a "tight" letter that is "fit for publication"), it can be useful. I have posted in the region of 100 online comments to articles but had six letters to the editor published.

So, for example, in the case I mentioned above, if this journal had online comments, I could have put a question out there enquiring about the follow-up study. The authors may or may not have replied (perhaps it would be good if there was pressure on authors to reply to such questions) but even if they didn't reply, attention would be drawn to the issue and perhaps somebody in a better position to follow it up might do so in some way at some point in the future.

No competing interests declared.


tkindlon replied to tkindlon on 08 Feb 2010 at 23:06 GMT

Apologies for typo in heading. It should be, "Registering reviews/meta-analyses is a good idea", of course.

No competing interests declared.