by Carl V Phillips
[Before praising a document whose primary intent is to keep menthol cigarettes legal, it seems worthwhile to start with clarification and disclaimer: As those of you familiar with the NGO that publishes this blog know, CASAA does not take positions on the regulations of cigarettes. As those of you familiar with my work know, a lot of my current research is about modeling consumer responses to changes in the tobacco market, and as a few of you know, one of my results is about the impact of a menthol ban on the black market and thus on overall consumption. This post is not about favoring any particular approach to cigarette regulation nor related to my modeling except to the extent that it covers some of the same ground.]
I suspect that few of you have seen Lorillard’s response to the FDA Center for Tobacco Products filings and request for comments that relate to a possible ban of menthol flavored cigarettes. But if you have any interest in legalistic scientific documents (or is that scientific legal documents? I am not quite sure what to call these) you are going to want to read it. It is really a tour de force as such things go. If there is such a thing as true art in that genre, this is it. (Links here under 10.21.2013; the ANPRM document is shorter, summarizing details from the other, so it gets to the interesting points quicker and it is more readable.)
The most notable bit of content, from where I sit and for the purposes of THR (see! this is not off topic), is the observation that FDA is not holding itself to the standards it imposes on others, particularly regarding MRTP filings. The MRTP process is the path that theoretically allows tobacco product manufactures to make claims that a particular product is lower risk than another, as well as a path to introducing new products that meet the standard of being lower risk. I say “theoretically” because the standards demanded for evidence might be so onerous that nothing can ever meet them. Some of those standards are clearly valid, like clearly stated methodology and reproducibility of the analysis. Others are naive, simplistic, or wrong, like blindly favoring some types of studies over others. Lorillard shows convincingly that FDA fails to adhere to these standards (as well as various standards of good science).
Lorillard argues that when the same phrase is used in multiple places in a piece of legislation — the phrase in question being “scientific evidence” — it must mean the same thing. FDA has created a very high bar for that in the MRTP process but, the argument goes, is badly falling short of that in issuing their analysis of menthol cigarettes. This creates a very interesting situation. Of course, the easy way out is they might just fold on the menthol point and not do anything that would force this contradiction to be litigated. The best outcome for consumers and public health, though, would be for them to concede (or be forced to yield) that the standard is far too restrictive and legalistic, and effectively prohibits doing science. Science is a process of trying to figure out the truth as best we can, in contrast with some legalistic processes, which are more about dotting all the i’s and jumping through hoops.
A central argument is about FDA’s conclusions that menthol trialing increase the probability of becoming a smoker compared to trialing non-menthol cigarettes, and as well that menthol increases “nicotine dependence” and reduces success of smoking cessation. Lorillard is clearly right that the studies cited in support of the first of these claims have important flaws. They have been used to try to quantify the claimed increase, but are clearly not adequate to do that. In particular, the studies seem to be claiming (to the extent that they can be interpreted based on the paltry bit that is published — a great deal of what would be needed to make sense of them has not been published) that all association observed is causation, which is absurd.
But here is the thing about the main claim that FDA wants to make: Anything that lowers the quality of
a this tobacco product category for some users (without increasing it for others) will tend to discourage some people from starting to smoke and encourage some people to quit. (The question about “nicotine dependence” is rather different.) If menthol flavored cigarettes were caused to not exist, there are some people — characterized by (a) close to indifferent about preferring smoking or not smoking and (b) like menthol — who will be tipped from smoking to not smoking. Of course, since more people dislike menthol than like it, if menthol flavoring were required for all cigarettes, there would be an even greater effect.
FDA faces two problems in employing such scientific reasoning, though: First, as Lorillard points out, they basically preclude it and demand only regimented forms of data collection — no matter how much less informative that might be than other sources of knowledge — rather than actual scientific analysis. Second, the logic requires acknowledging that people smoke because they like it, some people use menthol because they like it, people quit because they prefer to quit, and such. This runs directly contrary to the tobacco control industry’s notions of why people smoke (and why they use menthol), which is basically they are possessed by demons. I actually think that FDA is coming to grips with the fact that they cannot continue to endorse the demonic possession fiction and make defensible policy, but they are not quite ready to step up yet.
You might have noticed my tortured phrasing, “If menthol flavored cigarettes were caused to not exist”. This was a recognition of the fact that a ban would not actually cause this. Lorillard also argues, at length and quite convincingly, that banning menthol cigarettes would cause a huge boost in black market sales, and this would have all kinds of bad consequences. The problem with the black market for society (as opposed to the problem it poses for manufacturers and governments, taking away their profits) summarizes to the truism (my words, not Lorillard’s) that once you ban something, it is impossible to regulate it. In this case, it means that quality probably decreases (possibly including greater health risk), price decreases (due to lack of tax payments), and the merchants have no reason to avoid selling to minors.
Circling back to the implications for THR-related policy, Lorillard is arguing that all consequences need to be considered. For MRTP applications, manufacturers are required to assess the effects not just on those who follow the script (switching to a lower risk product) but those who do not (non-users who avoid high-risk products but are inclined to use the low risk product), and show there is nothing that would be considered a problem (though there is no standard for assessing that). It is a standard that has everyone scratching their head about how to do the assessment. If FDA is held to this standard for banning menthol, it is difficult to see how they could possibly meet the burden.
Further on this theme, Lorillard argues out that FDA mysteriously changes back to a “we just don’t know enough” attitude on the question of whether consumers perceive menthol cigarettes as being lower risk. Lorillard points out that the evidence supports the conclusion that there is no such perception (and, indeed, that menthol is perceived as being more harmful), much more solid evidence than supports some of the other claims that FDA declared to be true.
The good news would be for FDA to see the light about their unmeetable standards of evidence, and realize that when they try to do science they do not follow simplistic rigid rules like those, and so relax the MRTP application burden. The entire Lorillard filing, even though it primarily pursues different mission, is a massive indictment of the CTP’s current relationship with scientific inquiry, as it exists on paper, that effectively prohibits action of any sort. Unfortunately, experience suggests they will just back off on menthol to avoid the fight.
One additional point: There is a great subplot about peer review built into this. Peer review is correctly interpreted as the process by which scientists etc. get the input of others to make their work better, and thus relies on the assumption that everyone is in pursuit of the truth. The term is often debased and used to refer to a gatekeeping process for journals that in public health and many other fields is more about politics than real peer review. But whichever interpretation, Lorillard makes the observation, in effect, that FDA only pretended to have their material peer reviewed: They collected comments, but when some of those comments pointed out fundamental problems with their analysis they neither attempted to take advantage of the advice to improve their work, nor treated the criticism as gatekeeping and removed the bits that the reviewers objected to. This is just one more reason why the phrase “this is peer reviewed” is close to uninformative about scientific quality without further clarification.
Finally, I will note that there are some points in the Lorillard analysis which I disagree with. But this post is already too long to go into those. Besides, pointing those out is not my job (unless someone wants to hire me to do it :-).