Mainstream science reporting

The Washington Post’s outgoing ombudsperson, Deborah Howell, has an editorial about science reporting. She responds to reader and professional reactions to a prominent news report last month, reporting that statin drugs may reduce the incidence of heart attacks by 44%, even in people with no history of heart disease or stroke.

The basic problem, pointed out by former NIH chief Harold Varmus among others, is that a 44% reduction in a very small value (1.36%) is not very much of an effect. In a different way of looking at the problem, “nearly 97 percent of the people using the drug would not see any benefit”.

Howell focuses much of her editorial on the underreporting of corporate funding for medical studies – truly a concern, but one that often hides an agenda to supplant industry funding with federal money. (I figure, why not both?)

But in a much more interesting passage, she quotes the Post’s science editor, who ultimately makes the decisions about what to report and how to place it:

Nils Bruzelius, The Post's science editor, said, "I thought the story and Page 1 play were justified because the potential impact was significant, even as I understand the criticisms. There's an inevitable tension between the desire of reporters and editors to get good play for their stories and the need to avoid hype or overstatement, and we feel this very acutely in dealing with scientific or medical stories, because the advances, even those that prove to be part of something very big, usually come in incremental steps. I've long believed that science and medical stories enter this competition at some disadvantage. I certainly don't have data on this but I suspect that most of the top editors who make the front-page decisions tend to be less drawn to these topics than the average reader because, with a few exceptions, they are a naturally self-selected group who got to where they are by dint of their interest and ability in covering such topics as politics, international relations, war and national security -- not science."

This is a revealing statement. I don’t think it’s unique to science, though – after all, most political processes are incremental, and involve even more arcane topics like parliamentary procedure, budget accounting, and obscure officials. A story about health care reform has to describe these kinds of things just the same as a story about personal genomics. If there is a difference, it is that the science topics receive vastly less coverage, so that there are few people following the incremental steps. That means that each of the infrequent stories must either contain much of the same background material, or else must be very superficial. At the other extreme is sports reporting – in which there actually are constant new results, and for which most readers will know the major teams, players, and rules of the game.

So the real problem isn’t the nature of the subject, it’s the nature of the editors – the last part of Bruzelius’ quote. They understand politics. They don’t understand science. No training in it. Little feeling for what is realistic and what is fantasy. And unlike politics – in which few reporters are afraid to editorialize – there’s little attempt to strike a consistent editorial position.

(via A Blog Around the Clock)