Lisa Askie: Timely results

HOW long should it take from the time a clinical trial is completed to publication of the results?

A recent review of clinical trials published in 2009 found it took about 2 years from completion of a trial to the results appearing in a journal. The authors concluded that this information supported US government initiatives that require the results of clinical trials involving approved or licensed medical devices and drugs to be reported within 12 months of completion of a trial.

What is the situation in Australia? The simple answer is we don’t know.

Australia does not have any legislative requirements for the mandatory registration or reporting of clinical trial information or results. There are some levers that encourage registration of a trial prior to enrolment of the first participant, including this being a requirement for later publication in the major medical journals.

Some (but not all) ethics committees require evidence of prospective trial registration before they assess or approve a trial. Most ethics committees and funding agencies request information about a trial’s final publication but there are few, if any, sanctions if researchers don’t comply with such requests.

The Australian New Zealand Clinical Trials Registry encourages researchers to update their trial’s record to include the completion date and resulting publications but, without any requirement to do so, very few comply (currently only 17% of completed Australian trials list this information).

Why is getting trial results published in a timely manner important? Does a failure to do so compromise clinical care?

It would seem self-evident that publishing the results of research as soon as is feasible is a scientific, ethical and moral imperative.

Scientifically, it allows others in the field to recalibrate the current state of knowledge with the new findings and plan future research in appropriate directions. Ethically and morally, researchers have a responsibility to their trial’s participants to ensure that the information willingly shared is added to the public record and thus furthers scientific knowledge.

When this does not happen, important findings (either positive or negative) are unable to directly impact on clinical care or public health policy in the way we would expect.

A recent example is the 7-year delay in the publication of a huge trial of vitamin A supplementation and deworming involving a million Indian preschoolers. The trial’s results have significant worldwide policy implications yet the reasons why their publication was delayed remain unclear.

If publishing trial results in a timely way is important, what is preventing researchers from doing so?

There are several possible issues. For large trials with complex data collection from multiple sites, it takes considerable time to collate, check and “clean” the compiled data. To ensure high-quality data, these processes cannot — and should not — be rushed.

For trials without good, supportive infrastructure, this can delay analysis and subsequent publication.

It is known that it often requires sequential submissions to multiple journals before a trial’s results are accepted for publication. And it has been empirically demonstrated that trials with “positive” results get to publication significantly earlier than those with negative results, perhaps reflecting journal editors’ greater enthusiasm for more “exciting” findings.

In recent months, there have been strong calls for all data from all trials to be made publicly available as soon as possible. While this would undoubtedly improve the timely availability of trial results, there may be unintended consequences of full open access to all trial data, such as the risk of spurious results from non-peered reviewed, selectively reported, data dredging.

Getting trial results published as soon as possible and thus accessible to clinicians, policymakers and the public is a highly worthwhile goal, but this should not be done at the expense of scientific quality and integrity.
 

Associate Professor Lisa Askie is a principal research fellow in medicine with the NHMRC Clinical Trials Centre, Sydney Medical School, University of Sydney.