Wednesday, November 30, 2005

When You Don't Like What You Hear, Attack the Messenger - But Don't Discuss the Issues

In response to my post yesterday which provided a detailed analysis questioning the claim that smoking bans lead to immediate and drastic reductions in heart attacks, a prominent anti-smoking advocate publicly attacked me as follows:

"My general approach has been to ignore Michael Siegel, but his latest "analysis" is just too strange. He is publicising a "study" on the "Smokers' Club" web site. This "study" does not even account for population growth and there are no controls ... but only Mike can tell real quality."

The Rest of the Story

First of all, had the advocate taken the time to actually read and think about what I wrote instead of simply publicly attacking me, he or she would have noted that I was not relying upon the Smokers' Club web site study, but that I analyzed the data myself. It is my analysis (and not just what Michael McFadden and David Kuneman did) that suggests that there has been no observable (or at least no substantial) decline in heart attack admissions due to state-wide smoking bans. Even if I had relied upon the Kuneman-McFadden study without performing my own analysis, I would argue that it isn't invalidated because of who the authors are, but that it needs to be examined based on the actual data it presents. But this analysis is not going to be shot down on the basis of its authors being affiliated with the Smoker's Club, because I stand by the analysis, I think it is entirely an appropriate one (certainly with my extensions that extended the baseline period and used two comparison groups), and I think it demonstrates that there simply was not a dramatic decline in hospital admissions following the implementation of state-wide smoking bans.

Second of all, there is simply no way that population growth can account for the failure to observe a 27% to 40% decline in heart attack admissions (or even half of that) if one existed. Plus, if population growth were explaining the failure to detect such an effect, then there would have had to be basically at least a 13% difference in population growth between the smoking ban and comparison states and that difference would have had to be systematically different between the two groups.

Third of all, had the advocate read and thought about my comments before attacking me, he or she would have noted that I did employ 2 comparison groups - first, all of the states in the HCUP database in which a smoking ban was not implemented during the study period; and second, the entire country. If the entire country does not qualify as a comparison group, then I don't know what does.

Look - I am not claiming that this analysis is perfect or even that it suggests that smoking bans do not have a significant effect on heart attacks. My only point is that I think it is simply premature to be publicly claiming that smoking bans will produce a 27%-40% reduction in heart attack admissions. That's it.

Most importantly, I think I am bringing to light an important issue that deserves careful attention and further scientific analysis. I think that it deserves consideration. But attacking and insulting the person who makes such a suggestion without even attempting to read the entire piece or to discuss the scientific issues intelligently is just not appropriate.

And what it suggests to me, honestly, is that there is something missing. And what I think is missing is a sincere interest in discussing the science and in considering the different possibilities for what might explain the observed findings in the existing literature. This is precisely why it suggests to me that for some in the anti-smoking movement, the agenda seems to be driving the interpretation of the science rather than the science driving the agenda.

This has been a true learning experience for me. Never did I dream that some day, after 21 years of experience in tobacco policy research, as a statistical editor of perhaps the top tobacco control journal, and with over 50 peer-reviewed publications in top public health and medical journals, I would present a reasonably detailed scientific analysis of a tobacco control policy issue and then be publicly attacked and insulted for having the courage to present my opinions.

But because they apparently go against the anti-smoking agenda, I have now been publicly attacked and insulted. That's a shame.

But what's even more of a shame is that this issue really has nothing to do with the anti-smoking agenda. The issue of smoking bans should be decided based on the evidence regarding the health effects of secondhand smoke on bar and restaurant workers. It really doesn't matter, in my mind, whether smoking bans reduce heart attacks or not. The justification for such bans does not rest on such a determination. What does seem to depend on it, however, is the political ammunition that some anti-smoking groups and advocates seem to desire.

Finally, there are a lot of people who have expressed their scientific opinions on issues in tobacco control who I disagree with. But I would never attack and insult any of them, privately or publicly, for expressing their views. I think there has got to be a better way to practice public health and tobacco control.

No comments: