Friday, May 29, 2009

IN MY VIEW: Why Has the Science in Tobacco Control Deteriorated So Much? Because the Tobacco Industry Has Relinquished its Watchdog Role

Yesterday's story about the new study which, based on a cross-sectional correlation between greater enforcement of youth access to tobacco laws and youth smoking rates, concluded that there was a causal relationship, got me thinking about the reason for the deterioration of much of the science in tobacco control that I have been observing during the time I have been writing this blog.

And I think I've uncovered the primary reason for this deterioration: the tobacco companies have relinquished their role as watchdogs over the anti-smoking movement and its scientific claims.

To the best that I can pinpoint it, this shift occurred some time around 2000. Coincident with the tobacco companies' acknowledgment of the health hazards of cigarette smoking, it appears to me that they also changed their strategy with regards to challenging the scientific pronouncements of anti-smoking groups. Prior to that time, the tobacco companies would vigorously challenge the results of anti-smoking advocates' published studies. They would issue press releases, make public comments in newspaper articles, even take out advertisements challenging these conclusions.

But for about the last nine years, the tobacco industry has -- presumably as part of a concerted, strategic decision -- laid low and allowed the anti-smoking advocates' and groups' scientific claims to remain unchallenged. They have remained relatively quiet and are basically allowing the anti-smoking scientific claims to remain unchallenged in the public eye.

I'll tell you why I think this is the critical factor that has led to the deterioration of the quality of tobacco control science.

Back in the 1990's, any time anti-smoking researchers or groups would publish scientific papers, they would be very worried about the potential reaction of the tobacco companies. Prior to submitting any paper, researchers would consider the questions "What will the tobacco companies say?" and "How can we make sure we can defend these conclusions against potential tobacco industry criticism?" The same was true of conclusions disseminated by anti-smoking groups.

In fact, much of my role in the movement was to serve as a source for assessment of the strength of conclusions being made by anti-smoking groups before they went public with them. Every day, I would get calls from anti-smoking groups asking me to review their fact sheets to make sure they were solid and that the claims would be defensible against any potential tobacco industry attacks.

There was a pervasive sense of fear among all of us in tobacco control that if we slipped in the slightest, the tobacco companies would be there to attack us and publicly shoot down our statements.

Thus, we were extremely careful in drawing conclusions. There was a much higher burden of evidence required before causal conclusions were drawn. A single cross-sectional study would almost never be relied upon to draw a causal conclusion, for example.

Without the tobacco companies playing their watchdog role, however, there is no longer this sense of fear. There is no longer a concern about results being attacked or conclusions being challenged. It has become essentially a free-for-all, where anything goes and no one has to worry about their conclusions being challenged. In fact, the movement has evolved into one where challenging its conclusions is tantamount to working for the tobacco industry. Anyone who does challenge the conclusions of the movement is attacked and accused of being a tobacco industry mole or sympathizer. Thus, there is little threat that conclusions of anti-smoking researchers or groups will be challenged in the first place or that if challenged, the individual criticizing the conclusions will be taken seriously.

Back in the day, for example, we would never calculate population attributable risk based on the results of a single study. We would wait until multiple studies provided confirmation of a causal link and only then would we produce estimates of the total population burden of the exposure. Now, a single study suffices for researchers to make population estimates. Just yesterday, I discussed a study where based on a single cross-sectional study, a researcher estimated the total number of smokers prevented by youth access laws nationally. The American Legacy Foundation used a single study to estimate the total number of deaths prevented by the "truth" campaign. Other researchers used a single study to derive estimates of the number of youths who would be prevented from smoking by giving movies that depict smoking an "R" rating.

Ironically, the Rest of the Story is one of the only watchdogs there is to monitor the quality of science in tobacco control.

In the long run, I believe the tobacco companies made a wise decision. First, they help shed their public image as corporations that are undermining the scientific conclusions of the health community. Second, by letting the anti-smoking groups run rampant, they probably realize that these groups will eventually destroy their own credibility by making more and more outlandish statements, until finally, the public loses faith in the scientific integrity of these anti-smoking groups.

Ironically, the tobacco companies actually have a lot more legitimate criticisms of the validity of anti-smoking groups' scientific claims now than they did during the time when the companies were actually challenging these claims.

Another interesting aspect of this story is that I have realized that the peer review process is not necessarily adequate to protect against the publication of poor quality science. The tobacco industry actually played a much more effective role in safeguarding the quality of tobacco control science in the past than the peer review process does now.

No comments: