Thursday, September 27, 2007

The Dangers of Science by Press Release: Putative Effects of Scottish Smoking Ban on Heart Attacks Gets Widespread, Uncritical Media Coverage

An article published earlier this week in Spiegel Online informs the public that the smoking ban in Scotland resulted in a dramatic 17% decline in heart attacks during the first ten months that the ban was in effect. According to the Spiegel article, there were 551 heart attacks that failed to occur due to the smoking ban.

The article is based on unpublished data released via a press release by the University of Glasgow, home of the lead researcher on the study. To the best of my knowledge, the study itself is not easily available for public scrutiny.

In an uncritical recounting of the data and conclusions in the press release, Spiegel writes as follows:

"it seems a miracle has happened: The number of heart attacks in Scotland has suddenly dropped by no less than 17 percent in a single year. What has happened? Have the Scots stopped eating red meat? Has the whole country started knocking back cholesterol medication? Are they all training for the marathon? No. The reason is much simpler: Scots are having fewer heart attacks because they are no longer inhaling other people's cigarette smoke when they sit in the pub, the train or the office. Scientists at the University of Glasgow reported last week that things have become remarkably quiet in the country's heart clinics since smoking in public was banned in Scotland in March 2006. In nine selected Scottish clinics, 3,235 heart attack victims were brought in during the 10 months before the ban. In comparison, the number for the 10 months after the ban was only 2,684. Now the Scots and the rest of the world are marvelling at 551 heart attacks that never happened -- simply because of cleaner air."

The Rest of the Story

While Spiegel Online may be reporting that there were 551 fewer heart attacks in Scotland due to the smoking ban, Siegel Online is reporting that this conclusion is far too premature, and that the simple fact that there was a reduction in heart attacks from one year to the next is not sufficient evidence to conclude that this reduction is attributable to the implementation of the smoking ban.

In fact, I presented evidence that a 17% drop in heart attack admissions occurred in the 3 months prior to the smoking ban -- obviously, that much more drastic decline in heart attacks was not attributable to the smoking ban. Clearly, it is premature to jump to these conclusions.

This story demonstrates the danger of the science by press release approach. This is precisely why it is not prudent to issue findings that have not been validated by a peer review process. Suppose that upon peer review, it turns out that the study findings are unwarranted and invalid. Are the researchers going to then issue a press release stating that they were wrong, that their conclusion was premature and invalid, and that all news articles that the public throughout the world read to the contrary should be disregarded?

I highly doubt it.

Should these findings be invalidated based on peer review, it will at this point be inconsequential. The point is that the word is already out there. The conclusions have already been disseminated. It is too late to retract them if they are indeed wrong.

I view it to be inappropriate and perhaps unethical to release the results of a scientific study prior to peer review and publication. There is one exception, and that is IF (and only if) one makes the study available for public scrutiny and review. If you are going to go to the press to disseminate study findings, then you have to be willing to allow the public to see the study upon which those findings are based. Otherwise, there is no opportunity for review and scrutiny of your work.

To see how science by press release subverts scientific integrity, suppose that RJ Reynolds were to come out with a press release stating that Camel No. 9 cigarettes were found to reduce cancer risk by 20%. Newspaper headlines widely carried the headlines that Reynolds has produced a safer cigarette.

The first thing that anti-smoking groups would want to know is whether or not this research had been peer-reviewed and published. Suppose that it had not, and that the company did not release the study upon which these conclusions were based. There is no doubt that anti-smoking groups would condemn RJ Reynolds, question the validity of the study results, and claim that the company was widely misleading the public. RJ Reynolds would be attacked for violating established scientific standards.

Well if those standards apply to RJ Reynolds, then why don't they also apply to research which draws conclusions that are "favorable" to the anti-smoking cause?

Are we no longer required to adhere to solid scientific practice in tobacco control?

Apparently, we are not. Somehow, I must have missed that memo.

No comments: