Wednesday, May 18, 2005
That's the headline of an editorial in the May 17 issue of PLoS: Medicine, an open-source medical journal, whose editors have themselves stated that they will not become “part of the cycle of dependency…between journals and the pharmaceutical industry.” It has been clear to outside observers for some time that the current publication system was failing, just how is only slowly coming into focus. How can we, as readers, know when a drug trial, published in the vaunted JAMA, is really as valid as it seems? I've certainly gotten excited and posted about a few such articles on this blog. Should I retract?
Smith's ending point, his solution to this mess, leaves me disappointed. He suggests that journals should stop publishing trials altogether, and "concentrate on critically describing them." I'm not sure what this means, really, or how it could work (would we rather have trials just posted on the drug company's web site, trusting that it's valid?), or even why this it would be a solution. Unfortunately, I don't have a better suggestion.