Journalists (and scientists) should use absolute risks

Posted By


Stories like today’s about naproxen causing “a 50 percent greater risk of heart attacks and stroke than placebo” can be meaningless if they don’t provide the ABSOLUTE risk. The 50% figure is the relative risk — naproxen’s rate relative to placebo. But we’re not told the ABSOLUTE rate: how many people actually had heart attacks and strokes in each group. The difference could be very small if the ABSOLUTE numbers are small.

But it’s not just journalists who fail to give the absolute risk figures. The New York Times reports that those making the naproxen announcement yesterday would only say that 70 people out of 2,500 in the study experienced heart attacks and strokes. They would not give the numbers for each group. You can do the math to figure it out, but consumers shouldn’t have to do the math. Government officials and journalists should help.

Mea Culpa Disclosure: I violated the absolute risk tenet myself in my Dec. 17 entry when I wrote, “Pfizer has stopped a clinical trial for its blockbuster drug Celebrex because it was shown to cause 2.5 times more heart attacks than did placebos.”

You might also like


Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.


January 3, 2005 at 10:57 am

Do you think that placebo drug trials have a place in medical research–especially in Western nations where the “standard of care” is generally NOT “no care” (i.e., placebo), but some previous type of care? I am not a public health or medical person, but was introduced to this issue in a bioethics class and have continued to be fascinated by it.