5th World Conference of Science Journalists

PRESENTER: Chris Del Mar
SESSION REPORT: Medicine breakthroughs: The Clayton’s Promise

By Andreas von Bubnoff

Why is it that medical studies that are widely reported in the media sometimes later turn out to be wrong? One reason is that there are many different designs of medical studies, which means that the quality of the evidence is different as well.

In this workshop on Controversies in evidence-based medicine, Chris Del Mar, Dean of Health Sciences and Medicine at Bond University in Queensland, used many examples to explain how different study designs affect the reliability of medical studies.

For example, he said, a randomized controlled clinical trial found that beta-carotene can increase the risk for cardio-vascular disease. The trial overturned the results of previous observational studies, most of which had suggested a benefit of beta-carotene in the diet.

Randomized clinical trials, or RCTs, are more reliable than observational studies, because they get rid of certain biases such as differences between people who choose to eat beta-carotene from people who don’t. Such people may be different in ways other than taking beta-carotene.

One useful source of information on RCTs, Del Mar said, is the so-called Cochrane Collaboration, which pools the results of several clinical trials.

The results of such pooling of trials can sometimes even change clinical practice, he said. “I am amazed that you guys haven’t tapped into that much more,” he said. “There is more information on RCTs in there than in any other database in the world.” Unfortunately, he added, free access to the data is only available in some countries where governments are paying for a national subscription such as in Australia, but not in others such most of the U.S.

Del Mar also explained when a study is considered statistically significant. He said that even statistically significant results can sometimes be meaningless if the effect is below the clinically important threshold.