Media reporting of medical studies isn’t broken, but it does have a major crack in it. Fortunately, it would be easy enough to fix, if we can convince those who write about medical studies to stray from the current template.

Here’s the template, courtesy of a recent Vancouver Sun article. The study may not be familiar to you, but the shape of the article will be.

Start with an accurate, but not overly sensational headline. In this case: Alta. researcher links moles, freckles to eye cancer.

Hit the nub of the study in the first couple of grafs:

People who have freckles and moles have a higher chance of developing eye cancer, according to a Canadian study published in the U.S. medical journal Ophthalmology this month.

Edmonton ophthalmologist Dr. Ezekiel Weis says he and his team from the University of Alberta have completed the world’s first conclusive study into the connection between skin blemishes and cancer of the iris, also known as Uveal cancer.

Weis’s analysis found that people with many freckles or moles on their skin have between two and four times the chance of developing melanoma of the eye than people with no such blemishes.

Stir in a little advice from the study team.

Weis said Monday that doctors need to be vigilant with patients who have many moles or freckles and warn them to try to stay out of the sun.

And then, way down in the story, hint at — but don’t spell out — the real-life implications of the study:

Uveal cancer is extremely rare, affecting only about six out of a million people annually… Once contracted, the mortality rate of Uveal cancer patients is 50 per cent.

(Note: This isn’t the whole story, but it does accurately describe it.)

It’s not until I, a many-freckled person, get to the very end of the story that I see that the probability that I will die from uveal cancer is not the three-in-a-million for unfreckled folk, but instead six- or 12-in-a-million. (Multiply the number of cases by the mortality rate.) I should stay out of the sun because there is, at worst, a 0.0012 per cent chance I will die of uveal cancer

That’s the problem. We are fed a steady stream of media reports on medical studies that lead with big numbers (the 2-4 per cent increase) that hide what, in most cases, is a tiny real-life impact. It is rare to see that real-life implication spelled out.

(The argument that the information is there doesn’t really stand up, given what we know about the number of people who make it all the way to the end of newspaper stories. It’s also not nearly as clearly spelled-out as it would have been had the lede been, say, “If you have freckles or moles, you have a slightly-greater-than miniscule chance of getting uveal cancer.”)

It’s easy enough for reporters (or their editors) to fix: add a high-up graf, a sidebar or info box that tells us what study really means. Simpler still would be to require any reporter assigned to cover the release of a medical study to ask the question, What’s the real-life implication of this for my readers?, and feature the answer prominently.


Tags: , ,