You might think the New York Times has created a special medical news section if you open to pages A12 and A13 today. The paper runs three interesting medical stories, including a swine-flu follow-up worth noting here, especially because I posted on the coverage yesterday.
The CDC, according to a piece by Donald G. McNeil Jr., is casting doubt on Monday's frightening swine flu predictions--30,000 to 90,000 deaths, 1.8 million hospitalizations, and so on. In addition to quoting a CDC official on the record, he adds comment from a Harvard epidemiologist who notes that the epidemic seems to be fading in the Southern Hemisphere. Seeing this in print, it's clear that everybody should have raised questions about those fantastic numbers--as I should have in yesterday's post. A sharp piece of reporting.
On the next page, Gina Kolata reports on a JAMA study this week finding a steady drop in rates of hip fractures over the past two decades in Canada. She includes a thorough discussion of why this might be happening, and concludes that it is "a medical mystery."
She writes that the osteoporosis drugs called bisphosphonates, which became available in the mid-1990s, do not explain the decline. She does not, however, tell readers what those drugs are. I would suggest that most people taking Boniva or Fosamax do not know they are taking a bisphosphonate. We should give our readers the brand names they are familiar with, even though the medical journals don't.
And Roni Caryn Rabin rounds out the trio of medical stories with a problematic piece on a study in Cancer Research showing that tamoxifen (one brand name is Nolvadex, but this one is probably better known by its generic name) increased the odds that long-term users will develop a rare but aggressive new breast tumor. Several experts caution that the benefits of tamoxifen outweigh the risks.
Rabin, unfortunately, leaves out the absolute risk figures. She says long-term tamoxifen users were "possibly four times as likely as nonusers" to develop the new tumor. Meaning a 4 percent risk, compared to 1 percent? Or 40 percent compared to 10 percent? The figures are missing from the press release and the abstract, too.
Rabin does say the number of women who developed the unusual tumor in women who took tamoxifen for five years or more was small. And the risk was not found in women who took tamoxifen for one to four years. Both of those points mean "the finding of a four-fold increase was questionable."
Here's the judgment call: We don't want to hide this kind of news from our readers, and they would be unhappy with us if we did. But if the finding was "questionable," is it worth a story? And should it have a headline that reads, "Rare Side Effect Is Seen in Long-Term Use of a Breast Cancer Drug"? That hed doesn't sound questionable to me.
And, I've saved the best stuff for the kicker: Thanks to Gary Schwitzer and Ivan Oransky, who've been calling attention to this on the web, I found Dr. Len's Cancer Blog, written by Dr. Leonard Lichtenfeld, Deputy Chief Medical Officer for the national office of the American Cancer Society.
He doesn't solve my absolute risk problem, but he notes that the increased risk reported in the study was a range--between 3 percent and 1900 percent! (See his post for further explanation.) And similar figures had been reported by the same researchers in 2001.
The title of Dr. Len's post? "When the 'News' Isn't News."
- Paul Raeburn