Most Media Coverage of Drugs Highly Biased

A study of how the mainstream mass media covers health found that many news
stories on drugs fail to report side effects or researchers' financial ties
to the companies that make the medications. The researchers looked at 207
newspaper and TV stories from 1994 to 1998 on three drugs: aspirin; Zocor,
a cholesterol-lowering drug; and Fosamax, an osteoporosis drug.

In the 170 stories that cited experts or scientific studies, half included
at least one expert or study with financial ties to the drug's
manufacturer. Of those, only 40 percent reported the potential conflict of
interest. The study also found that fewer than half the news stories
reported the drugs' side effects and only 30 percent noted their cost.

This report was published in the New England Journal of Medicine, whose
incoming editor has been charged by the FDA for an apparent conflict of
interest involving a drug company. He has admitted that he may have made a
mistake last year when he praised a new asthma drug made by a company that
had hired him to evaluate studies about the medication.

Additionally, forty percent of the stories studied did not report the
numbers behind the claims of medical benefits. Also, 83 percent of the
studies reported only the relative benefit, 2 percent reported only the
absolute benefit, and only 15 percent reported both.

For example, many 1996 stories about a Fosamax study said the drug would
cut an osteoporosis patient's risk of a broken hip in half - the relative
benefit. But most failed to include the absolute reduction in risk, from a
2 percent chance of a hip fracture to 1 percent.

Reporting only the relative benefit is an approach that has been shown to
increase the enthusiasm of doctors and patients for long-term preventive
treatments and that could be viewed as potentially misleading. In addition,
while most of the top medical journals require researchers to report their
financial ties to drug companies, some studies do not include the
information because a researcher fails to disclose it.

New England Journal of Medicine   2000; 342: 1668-1671.