I'm drooling, but I shouldn't be. John Carey of BusinessWeek has the cover story this week, entitled, "Do Cholesterol Drugs Do Any Good?" I'm not going to post an excerpt because I want everyone to read the entire article. It's that good - and that rare. I'm drooling over it but I shouldn't be, because the themes he explores should be part of all news stories making claims about new ideas in health care.
Carey explores absolute vs. relative risk statistics - showing how the latter are abused and are misleading to the general public to trump up drug benefits. But he also is one of the only journalists I've ever seen explore and explain another statistical tool - the Number Needed to Treat or NNT.
See his insert on this important statistic. It will raise questions about the way drugs are promoted and sold to hundreds of thousands of people who won't benefit from their use.
When I wrote the Statement of Principles of the Association of Health Care Journalists, I urged health care journalists to "consider explaining the 'number needed to treat' - the number of people you would have to treat with the experimental intervention (compared with the control) to prevent one event." And I included tips on how to do that. But it rarely, VERY rarely ever happens.
Carey and BusinessWeek hit a home run with this cover story. Kudos to them. Keep up the good work. Let's hope other journalists - and certainly consumers of news and health care - learn from this excellent piece of work.
Posted by schwitz at January 25, 2008 05:00 PM | TrackBackThanks for pointing out this piece.
It reminded me of an exchange I had with a group of physicians at a conference on evidence-based medicine. They were entirely focused on boosting the statin use among their high-risk patients.
When I commented that of course they couldn't promise an individual patient that he (the case being discussed was a man) would definitely benefit, they were befuddled. They simply hadn't wrapped their heads around the fact that even in a high-risk group, where the population benefit of statin use is well-documented (wiht NNTs better than 10), the overwhelmingly majority of individual patients will not alter their fate.
And as you say... it's too bad such reporting isn't run-of-the-mill.
Posted by: Andrew Holtz at January 29, 2008 04:01 PM