When Registering Trials Is No Help

The science journals jumped all over the "hidden data from clinical trials" issue this week with editorials or commentaries in the British Medical Journal, Nature, and Science. The trigger was a meta-analysis published in PLoS Medicine suggesting antidepressants in most cases are not much better than placebo.

The authors had to use the Freedom of Information Act to get results of trials they needed to conclude the analysis, and even then were denied crucial studies. The subheads said it all: "Highlights the ongoing problem of how study results may be distorted by failure to make data fully available"; and "Highlights the need for data to be transparent -- and for a mandatory database of all clinical trials."

Congress just passed a law requiring mandatory registration of all clinical trials and the posting of results. But it falls far short of what is necessary. It is not going to solve problems caused by the fact that industry still funds 70 percent of all trials. They get to decide what gets studied and how it gets studied because they foot the bill.

As Science pointed out is its inimitably circumspect way, the new law raises "further issues" that "require examination."

For instance, does the new law protect against a clinical trial not having an appropriate comparator drug instead of just placebo or an inappropriate comparator drug (like a low-dose of a less effective medicine that will no doubt make your newer drug look powerful, indeed)? No, it doesn't.

Does the new law protect against data fraud, like what happened in the Ketek case? No it doesn't.

Does it protect against insufficient informed consent for patients, especially in Phase I clinical trials? No it doesn't. Remember what happened in Great Britain when they tried a brand new biologic for the first time in humans in eight patients simultaneously and wound up killing one and almost killing the rest? Moreover, by not requiring registration of those Phase I safety trials and publishing results, the new law does nothing to warn other companies off dangerous classes of drugs, or unproductive pathways for their increasingly unproductive research-and-development labs.

But the reality is that the anger being expressed now about failure to disclose data in clinical trials should largely become a thing of the past if, and this is the big if, the Food and Drug Administration writes good rules and properly polices the new law. Let's hope that the medical journals and their news sections follow implementation of the new law as closely as they are following the old stories that resulted from the lack of proper registration and posting of results.

Posted by gooznews at March 7, 2008 08:48 AM