Sunday, November 15, 2009

Teach the controversy

Howard Markel, a physician who was interviewed for that Atlantic article on flu vaccines, wrote a letter to the editors clarifying his views and affirming that vaccinations are important for fighting the disease. The writers respond:
We appreciate Howard Markel’s desire to clarify his personal position on flu. As the article says, many experts believe flu vaccine is effective; some believe it is not. The point of the story is that we really don’t know one way or the other: our public health policy is built on a very thin base of evidence.
Wow, is that irresponsible. Thousands of studies show the vaccine to be effective, a handful do not. Therefore, "we really don't know one way or the other." By that standard, we have no idea whether germs carry disease, whether cigarettes cause cancer, or whether the Earth is flat or round. We should debate these things.

5 comments:

Unknown said...

And let's be clear. That irresponsibility will cost some people, and some peoples' children [how will they feel then???] their lives.

Steve Balboni said...

The cigarette.cancer analogy is really apt. Vaccines, climate change - how can we ever know? Contrarianism and skepticism run amok, aided and abetted by a media which has lost the forest for the trees.

Joe said...

I am not quick to heap scorn on the Atlantic article. The gold standard of epidemiological research is the double-blind placebo controlled test. Most epidemiology doesn't meet that standard, and obviously none of the observational research does. In addition, it is standard for epidemiological research to ignore multiple-testing problems; they do not use Bonferroni or Sidak adjustments. Consequently they conclude things like "women who eat breakfast cereal have more male children" (I kid you not), get published, and this becomes the headline of tomorrow's edition. The only study I've seen that takes a look at this is by Ioannidis (JAMA 2005), who found that 5 of 6 highly cited observational studies were contradicted by subsequent studies. So when I see a journalist who isn't falling all over himself to breathlessly write up the latest "coffee cures lumbago" story, I'm sympathetic to the intent if not the outcome.

Seth Masket said...

I take your point, Joe, and the article in question is more thoughtful than most journalistic coverage of medical findings. But in the end, it's still an article in which a journalist is falling all over himself to breathlessly write up the latest "coffee doesn't cure lumbago" story. Or, in this case, vaccines don't prevent the flu. The author is in love with a particular finding and is ignoring a lot of counter-evidence and indeed the entire scientific method in reporting it.

lidzville said...

To extend on Seth's point, IMHO American assigning editors have fallen a bit too in love with stories that seek out a piece of conventional wisdom against to cut. Slate.com's tendency to take this to an extreme has had an influence on editors elsewhere, I suspect.