Sim and her colleagues reviewed the publication status of all 909 clinical trials that supported the 90 new drug approval applications approved by the U.S. Food and Drug Administration between 1995 and 2000. They found the most important trials determining efficacy, and those with statistically significant results and larger sample sizes, were more likely to be published.

Sim said the study found evidence of selective reporting of the trials. For example, Sim and colleagues report a pivotal trial in which a new drug works better than an old drug is more likely to be published than a trial in which the new drug does no better.

"This is a form of publication bias that may lead to an inappropriately favorable record in the medical literature of a drug's true risk-benefit profile relative to other standard therapies, and can lead to preferential prescribing of newer and more-expensive treatments," she said.

The research is detailed in the journal PLoS Medicine.