Tuesday, September 11, 2007

Science by Press Release: The New Tobacco Control Approach for Disseminating Junk Science

In addition to representing shoddy science, the Helena and Helena-like studies of changes in heart attack admissions following smoking bans seem to have introduced a new style of scientific "inquiry" into tobacco control: science by press release.

Instead of waiting until research is peer-reviewed and published in scientific journals, anti-smoking researchers now seem to be releasing their results (but not their full papers) to the media prior to peer review. Just this week, we learned of two new studies - one in Ireland and one in Scotland - which purportedly show that the implementation of smoking bans in those countries resulted in a dramatic decline in heart attack admissions.

It appears that the Helena study set the trend for this science by press release approach. An April 2003 press release announced that the Helena smoking ban resulted in a 60% decline in heart attack admissions within the first six months that the ordinance was in effect. This press release was apparently issued when the paper was first presented at a scientific conference, but before it had been peer reviewed and published. The paper was not published until April 2004, one full year after its preliminary conclusions had been disseminated via press release. As it turns out, the data in the initial press release was wrong. The decline in heart attacks was only 40%, not 60%.

After researchers noted a 27% decline in heart attacks in Pueblo following a smoking ban, the results of their study was also disseminated by press release prior to publication in any journal.

Now comes the Ireland study, the results of which were disseminated widely to the public last week through the media, via press release (see page 69 of the media guide). However, the study was merely presented at a conference. All that I could find was an abstract. Apparently, there was no paper available for public review, no peer review, no publication in any scientific journal.

Days later comes the Scotland study, the results of which again appear to be presented at a conference, but not peer reviewed or published, and the methodology of which does not appear to be available for public review.

The Rest of the Story

First, I would note that it appears to be the most shoddy scientific findings for which this science by press release approach is taken. This approach makes sense. Allow the results to have a maximum impact on the media and the public before they are peer reviewed. That way, if the results need to be toned down for publication, the damage has already been done.

This was certainly the case with the Helena study. The media - including this CBS news story - widely reported a dramatic 60% decline in heart attacks. That turned out to be wrong, but so what? The news had already been reported, and the damage was done.

Second, I have to say that I believe it is inappropriate, and perhaps unethical, to use this approach of science by press release. I firmly believe that the time to release scientific results to the public is when the paper is published in a peer-reviewed journal. If the results are not published in a journal, or if they need to be released prior to publication, then it is imperative that the researchers make the entire study easily available for public review. Otherwise, there is no way for the public to judge the validity of the results. The public becomes totally dependent upon the researchers' word.

I find this science by press release approach to be yet another disturbing development in the tobacco control movement. This, too, is in my view contributing toward the degradation of the scientific integrity of the movement.

No comments: