New Rules for Publishing Clinical Trials

Back in December of 2005 I mentioned the clinicaltrials.gov web site. This site, where many large multi-center clinical trials are registered online, was primarily intended to store information on research being sponsored by the National Institutes of Health.

That site, and several others like it, expanded considerably as a result of a 2005 policy implemented by the International Committee of Medical Journal Editors. The ICMJE, as it is known, is a very influential group of the world’s most prestigious journal editors. Much of the leading clinical research is published in these medical journals.

In 2005, the ICMJE established the policy that authors wishing to submit manuscripts for consideration in any of the ICMJE journals which described the results of clinical trials could only do so if the clinical trials were registered with one of the online clinical trial registries, such as clinicaltrials.gov. Since that policy was established, the number of clinical trials registered on clinicaltrials.gov more than tripled, to over 40,000 today.

The ICMJE was criticized for at least two major flaws in its policy. First, it did not cover the earliest kind of research, Phase I clinical trials, which are usually designed to test the safety of a new compound or device in a relatively small population of research participants. Second, because the policy required that results of the clinical trials also be published online, many manuscripts were turned down for publication because the results had, in effect, already been published (online).

Fast forward to June of 2007.

This month, the ICMJE updated its policy to include Phase I clinical trials. This means that, for any clinical trial in which participants are enrolled on or after July 1, 2008, even if they are only small trials designed to see how toxic a potential new drug is, they must have been registered online in order to be eligible for consideration as publications in any of the ICMJE journals.

Many physician-scientists who conduct clinical trials are interested in publishing their results in prestigious journals so they can progress in their academic fields. They will not be able to publish in ICMJE journals if the clinical trials they conduct have not been registered.

Why would they NOT want to register the trials in the first place? Such pressure usually comes from the drug company sponsoring a particular clinical trial. For strategic business reasons, companies do not want to give away any information on its research and development activities, especially so early in the process. Not only does it potentially tip the company's hand to competitors, investors may be scared away by published scientific reports of "failed" ideas.

The second perceived flaw in the ICMJE policy, the paradox created by requiring online publication of results, making those results less valuable for publication in journals, was also addressed. ICMJE journals will now treat results that are published in the same online registry in which the clinical trial itself was published as brief abstracts (brief being defined as less than 500 words) or as tables of data as though they had not been previously published at all. Therefore, the results, when submitted to the journals as fully developed scientific manuscripts, would be acceptable to the journals.

Yes, these represent baby steps. But they are baby steps forward and do represent movement toward greater transparency when it comes to testing compounds which will be used to treat people.

~Joe

Related Topics:

• WebMD Video: Special Report: Clinical Trials
• New Avandia Study Doesn't Prove Safety

Technorati Tags: clinical trials, ICJME, pharmaceutical

What the FDA Guidelines are Recommending

This is the second of a two-part blog inspired by recent recommendations by the Food and Drug Administration (FDA).

The FDA's recommendations, published as guidelines on April 9 in the Federal Register for public comment, are intended to streamline the adverse event reporting process somewhat.

The idea is for adverse events to be reported to IRBs only when they are known to present unanticipated problems for the clinical trial. In other words, they must not have been anticipated when the clinical trial was developed and they must truly present a problem of some consequence to the clinical trial.

The guidelines further recommend that adverse events be reported in context and/or with some analysis that will enable IRBs to determine what implications, if any, the reports have on the continuation of the clinical trial and the safety of the research participants.

Administrative streamlining is generally a good thing, provided a critical step is not missed or information is not omitted. And providing context for data is also generally good.

The problem here arises with the FDA's suggestion that the entity providing reports to the IRBs is the clinical trial sponsor. It is true that if all of the reporting of adverse events that is currently required of investigators to the sponsor is done, then the sponsor is in the best position to identify a particular adverse event as an unanticipated problem. But what if a sponsor chooses not to do so, either intentionally or unintentionally, due to its overriding interest in bringing the product being tested to market?

Companies sponsoring clinical trials of products they own are already under tremendous pressure to see that the trials are finished as quickly as possible, as cheaply as possible and with as few adverse events reported as possible. Asking them to put individually-reported adverse events into context for the IRBs would seem to be asking for trouble.

At a minimum, in addition to submitting reports of adverse events (unanticipated problems) to IRBs, clinical trial sponsors should be required to post all adverse events received from investigators.

This posting could be done easily enough on the World Wide Web, and be made accessible only to those involved in conducting and regulating the clinical trial. The data should remain available for the duration of the clinical trial and throughout the FDA review process. Data would only be taken down if and when the sponsor withdraws its application for FDA approval.

Such open disclosure of all adverse event data would show how the sponsor interpreted the data of which it was aware, perhaps preventing some of the bias (recognized or not) resulting from the sponsor's financial interest.

~Joe

Related Blogs:

• FDA's New Adverse Event Guidelines
• IRB Shopping

Related Topics:

• WebMD Video: Behind the Scenes of a Clinical Trial
• FDA Panels: Conflicts of Interest?

Technorati Tags: clinical trials, adverse events, FDA, IRBs, health and wellness