One-third of All Studies are Left Completely Unpublished

By August 11, 2010Uncategorized

Sometimes there are no adequate words one can use to explain something. Indeed, there are some revelations that baffle people and should cause them to raise a great many questions. One such revelation is the recent story that a great portion of clinical trials of pharmaceutical drugs are never published or are buried where they cannot be readily accessed.

According to the Annals of Internal Medicine, nearly one-third of all studies are left completely unpublished and inaccessible to anyone but the group that performed the study. This analysis included medicines such as antidepressants, antipsychotics and blood pressure medication — medicines that society at large has a bit of a vested interest in knowing about.

All clinical trials are required to be registered at clinicaltrials.gov, but this is not the same as publishing. The material is not present in the public database of medicine studies, PubMed. Thus, it creates a situation where groups can claim they performed a study without having to demonstrate evidence. Fortunately, the industry at large operates on an assumption that a study not put through PubMed isn’t relevant, but it still raises uncomfortable questions about why they aren’t going public in the first place.

This creates a situation known as publication bias. By observing the requirement that they publish to the less public site, pharma companies then can pick and choose which articles see the light of day. Thus more positive articles and studies are shown to the public at large, creating a more positive perception of drugs such as antidepressants than the medical studies might perhaps warrant.

Doctors need to very seriously consider this. They often learn about a drug long after the trials are finished, and at a time when trials may not be being currently performed. If this kind of publication bias is genuinely the norm, do doctors want to risk prescribing these potentially-compromised medicines to their patients?

—-