by Carl V Phillips
The FDA just concluded the meeting of their Tobacco Product Scientific Advisory Committee (TPSAC) to review the MRTP application by Swedish Match to change the incorrect warning labels on their smokeless tobacco products. They applied for removal of the warnings that say that the products cause diseases that they do not actually seem to cause (meaning: to a measurable degree, of course) and to replace them with a warning statement that says while no tobacco product is safe, these are substantially lower risk than smoking. (Background on that here, here, and here.) In other words, they were asking to be able to state something that is beyond doubt and not be forced to make claims that are not supported by the science.
So how do you think this extremely reasonable and clearly valid supplication to the FDA went?
(Note that after watching the hearings for two days, I am not sure I can form a coherent narrative about everything. So this post starts out with narrative and then switches to a series of more disjoint points. Some of this is insider talk that casual readers may not appreciate. Also, casual readers will not appreciate the length. Basically, this post was designed for the science and/or policy wonks, though I hope it is interesting enough for others who are seriously interested in the topic.)
Perhaps the most striking part of the whole event was the mere fact that the second day consisted of TPSAC voting on a series of questions that the FDA staff posed to them. Science-by-committee is generally a failed process in the first place. It combines the groupthink effects that occur in juries and other social processes, wherein a bad idea can pick up steam with every new comment, along with a lack of personal responsibility due to individuals thinking they are just one member of a group. There are clear examples of both of these in this case. Of course, committee do not always go that way — e.g., Richard Feynman, on the committee investigating the Challenger accident, actually did some science and forced the committee to face the truth, much to the annoyance of bye bureaucrats — but there are not exactly a lot of guys like Feynman on these committees.
If something is obviously true beyond any reasonable doubt, there is about an 80% chance one of these committees will declare it to be true, and thus merely a 20% chance they will declare something clearly false to be true. In between it appears to be about fifty-fifty, regardless of the evidence. Also, there is is about a 75% chance that they will declare their conclusion to be definitive, and this seems to be not at all correlated with whether their conclusion is correct.
But this is worse. It is as bad as al that when a committee meets, reads, contemplates, and iterates. But this committee got one day of briefings and then had one short day to vote on a dozen question. No time to contemplate or research. Not that they necessarily would: They had material they could have read in advance, but some of them clearly either did not do that (as evidenced by many of their questions) or perhaps just did not understand it. Perhaps we cannot expect their understanding of this material and the relevant science to be as great as, say, mine, or that of experts in the industry, but you would like to think that they would have been among the few thousand people in the world with the most expertise. For the most part, they were not. (Maybe they should have picked a random physicist to serve on the committee.)
The voting process was made worse by TPSAC’s refusal to addresses the questions posed to them, even when they were the right questions. The first three questions, about the products causing gum disease, tooth loss, and oral cancer (which the warning labels claim) were put to them in terms of “Does the evidence support the statement that these snus products pose a risk of….” The committee changed the wording of these, before voting on them, to “Does the evidence support that these snus products do not post risks…” That was when I found myself wondering, is there a scientist in the house?
Any scientist would understand that the original question is a proper scientific question about whether the evidence is sufficient to rule out the possibility there is no risk, to a decent level of confidence, which is what is (or should be) meant when someone claims there is a risk. The changed version demands the impossibility of evidence showing there is absolutely no risk. Everyone should remember from grade-school science lessons that scientific evidence can never support such a negative claim. There is always a possibility that a series of events that has never been studied could lead to the outcome. More important in this case, the imprecision of epidemiology means it is never possible to rule out a small risk (for any exposure possibly causing any outcome), even if you look for it in exactly the right place. That is why we never ask the question that way.
It turns out that there actually is one real scientist who understood this on TPSAC, Jim Swauger from Reynolds, who holds the non-voting seat reserved for a large tobacco company. He protested that this standard was unattainable, and that (my paraphrase) it was not a valid scientific question for a scientific committee to be addressing. He even pushed this again a few times, which was a bit surprising given the statement from Reynolds from the first day (see below) along with his frustration (evidenced by several other comments) with the recurring theme that there is something fundamentally different about the Swedish Match products. I guess he is just too good and honest a scientist to not voice this frustration. Paolo Boffetta, who was sitting ex officio on the committee also voiced this observation once. Boffetta obviously understands this, but apparently a career of being a part of science-by-committee meant that his shut-up-and-go-along urges defeated any inclination to defend good science, and he did not press the issue.
When the voting members of TPSAC voted on the gum disease and tooth loss questions, as amended, all eight supported the obviously correct answer that the evidence does not support the claim that there is no such risk. After each vote, they were required to state their reason for the vote because — as stated by Zeller — that was as much a part of the record that FDA would consider as the vote itself. (Those were often pretty amusing, since about half the time they consisted of someone saying something along the lines of, “I voted that there was not such evidence because there is not such evidence.” Imagine an exit poll when one of the possible answers to “Why did you vote for Obama?” was “because I wanted him to be elected”.) It turned out that when giving their explanations, not one of these supposed-scientists gave as the reason that the stated standard can never be met for anything, about anything, and therefore it was obviously not met in this case.
Think about this: Imagine a pharma company wanted to get nicotine patch approved as an MRTP, to be sold as a consumer product as a lower-risk alternative to smoking. The impossible standard the TPSAC imposed would mean that they would not be able to dispute a warning label that it caused tooth loss.
The impetus for the bad question falls squarely on government officials, though good scientists would never have allowed the external committee to go along. It was triggered by Timothy McAfee from CDC, ex officio on the committee, playing an epistemic shell game: He argued that if the standard for imposing a warning were to require evidence that a particular product — as FDA thinks about products (i.e., a particular SKU) — causes a disease it would never be met, and thus the assessment should be based on more general evidence. This is right, of course, up to that point. But then he cleverly switched up what he was saying and conned the voting members of the committee by claiming that since it was established that the product category (smokeless tobacco) does cause these diseases, you would need affirmative evidence that the particular product does not. Both halves of that argument are wrong. As Swauger noted, the evidence is pretty much the same across the category, and the claims were no established more generally. And as McAfee himself began his argument, there is no way to get solid information about an individual product.
Indeed, on the first day, McAfee conceded that there is not enough data to impose the existing labels, but asserted that there is also not enough to remove them. Huh???
Swauger pointed out, correctly, that the evidence does not actually support the assertions about the category as a whole (Reynolds had previously applied to make the same label changes across the category). But that did not matter because McAfee was voicing the message that FDA had been delivering from the start (more on the mood from FDA below): Because the warning labels exist on smokeless tobacco products, what they say must be true, and so the only basis for removing the warning label is proving it is not true for a particular product.
Can you imagine if FDA had been in charge of other government decisions and applied this standard? Should we allow gay marriage or interracial marriage? Well, it is banned so it must be harmful, and thus we need scientific evidence that it causes no harms before we allow it. Should we repeal Prohibition? It must be that Prohibition is a good policy, since it exists, so we need scientific proof that there there is no risk from doing so. Should we emancipate the slaves? Should we allow women to vote? Should we allow AIDS sufferers with only months to live to try experimental drugs? No, not unless there is scientific proof that the drug will not cause them harm. (Oh, wait, FDA actually did do that last one.)
The third vote on “evidence of no risk”, about oral cancer, got even more absurd. The deliberations (a charitable description) centered on cherrypicking one study or subgroup from one outlier study that showed an association, and suggesting this was the relevant evidence. Never mind the fact that you would not even need that — if every study produced a 1.0 relative risk, or if there were no studies at all, you could still never say that the evidence shows there is no risk. Swauger objected again that the standard is impossible to meet in the first place, and also that if you have six or seven epidemiology articles, it is pretty much inevitable that one of them will have a positive association even if there is no causation. This was ignored.
When the vote came, there were actually three yes votes. In explaining his yes vote, Rich O’Connor said that there is sufficient evidence to say “there is no robust effect” and Kurt Ribisi expressed similar sentiments. This is right, of course, but this was them refusing to respond to the bad question as asked, instead of taking the more useful step earlier of joining Swauger in objecting to the wording. By contrast, Tom Eissenberg justified his yes vote based on the claim that the data about those who use the products show no risk. This is also true, but is a major logical fail for the reasons noted — the data cannot actually support a yes vote on the question as asked. Gary Giovino abstained, saying there is good evidence that the relative risk is about 1, but he was “not ready to go there” to say the evidence showed there was no risk (um, yeah, exactly what evidence would ever make someone “go there”?).
The vote was tied, 3-3-2. The explanations for the other abstention and the no votes introduced even more absurdity. I circle back to that in the discussion of groupthink, below.
Stunningly, it was Scott Tomar — who was arguably the worst anti-THR liar in the late 1990s and early 2000s — sitting ex officio, who spoke up after the vote to ask whether, if the question had not been phrased in terms of trying to prove a negative, some votes might have been different. Presumably Giovino would have voted with O’Connor and Ribisi, since his opinion about the data was the same. But no said a word in response to Tomar’s excellent observation and question. O’Connor or Ribisi should have noted that they actually did vote based on the scientifically proper question, not the one asked. Giovino could have confirmed he would have voted with them. But they were silent, in a testament to the behavior of science committees (absent someone with Feynmann-level fortitude) to just go with the flow.
There was a fourth unanimous vote on a “well, duh?” question about whether the comparison to smoking should be based only on diseases caused by smoking or all relevant health outcomes. I mention this only because of the absurdity of it: There are no relevant health outcomes that the ANTZ have not tried to blame on smoking. Other than perhaps cutting your finger on a snus tin, I cannot even figure what they might add to that list. Every voting member save O’Connor, who aded some substance, “explained” their vote that we should consider all outcomes by saying, per the above, “I think we should consider all outcomes.”
The fifth vote was on the question of the new label that Swedish Match proposed (and that Reynolds had previously proposed). It asked whether the evidence supports the statement that the health effects for the user are “substantially lower” than from smoking. This is another “well, duh” question, except this time the politics and the obvious scientific answer pull in different directions. You can guess the rest. Giovini suggested it should say “much lower” because that is a shorter word (this was a reference to the endless discussion from the day before about the methods used in Swedish Match’s consumer research — see below — though this does not make it less silly). Thomas Novotny argued that the word should be removed entirely because the quantity is ambiguous (so when it comes to actual scientific questions, they ignore the wording; when it comes to rough heuristic popular communication, they obsess about it). Zeller intervened and asked them to not fiddle with the phrase because that was the phrase that was put to FDA by the applicant and they need to rule on it.
And so Eisenberg immediately started talking about use during pregnancy and said that the label ought to add a warning about that. They proceeded to start talking about what specific warning labels should be added to the product. Even FDA was fed up at this point, and David Ashley intervened again to say “stop it!; we are asking you a specific scientific question; answer that question.” (Needless to say, my paraphrase. Ashley is rather more diplomatic than that.) He also pointed out to them that the next question — in case that actually reading the agenda in advance was too much to ask of them — is about warning labels specifically, but this one was not.
So they instead bickered for a while about whether the phrase “health effects” should be used since all the effects are unhealthy. This is not actually true, of course, though apparently no one understood the science well enough to mention that. Finally they figured out that health effects does no mean positive health effects, and that they were being stupid. One of the industry representatives pointed out that there appeared to be some serious prejudice on the committee. Um, yeah.
They then proceeded to ignore FDA’s requests and decided to change the question to include the caveat that exclusive use of the products was lower risk. Zeller gave up on trying to herd the cats, and — with an obvious sigh that is uncharacteristic for someone normally so scripted — asked them to at least explain how that word affected their vote.
They then voted on whether snus use is lower risk than smoking, and four of the eight voted no despite how obviously wrong this is.
Their epic fail appears to mostly because they apparently could not understand, despite the reminders, what question they were answering. Yes, we are dealing with some really great scientists here. Once again, O’Connor (voting yes) was the only one who could actually tried to explain his position as was asked, saying that the evidence about partial substitution was less clear (though this was neither a clear response to Zeller’s request nor entirely accurate). The other yes votes were Bickel, Giovino, and Ribisi, who did not explain how “exclusively” affected their opinions, but instead inserted caveats similar to those that were given as the reasons for the no votes, which were off-point rhetoric.
Before discussing that, however, I want to circle back. Because at this point you might find yourself thinking, “the TPSAC is trying to come up with rhetorical excuses to shoot this application down, refusing to even recognize the validity of the obvious truths, but at least the CTP staff — who actually make the decision — seem to pushing back against that.” That is true, but at the end of the first day I found myself thinking, “CTP is really trying to undermine this; all that remains is whether TPSAC pushes back against this obvious bias and does not let them get away with it.”
The first day started with Swedish Match, mainly in the person of Lars-Erik Rutqvist, presenting their case. This seemed like a waste of time. Swedish Match filed their application with all its detail and supporting documents; they wrote a readable summary of what was contained in the filing (since no one was going to read it all); they wrote a summary of the summary. Everyone should have read that, right? So what is the point of an oral presentation that can only address a fraction of the summary of the summary? It takes up time and runs the risk of being a distraction. And it was.
For example, Rutqvist spend quite a bit of time presenting data about the use of snus for smoking cessation. McAfee (legitimately) pointed out that they were not seeking permission to make smoking cessation claims, so that is irrelevant. The presentation of their study of consumer reactions to the proposed warning labels went even worse. TPSAC members glommed on to the observation that the study methodology presented consumers with the text of the proposed labels without the word “WARNING:” in front of it, as appears on packages. This bogged down into ten minutes of talking about that and other minutia of the study design and possible minor improvements in it. Eventually Rutqvist responded with “there are lot of ways we could have done this differently, which might have been better but might not.” But the damage was done — the groupthink was already fixated on irrelevant points that were like a shiny object for them because they had been discussed at length.
(Aside: Rutqvist should really fire whoever prepped him for that meeting. There is no one who knows the material better than he, and he is at home with truth-seeking scientific discourse. But he was totally out of his element being hostilely cross-examined, basically what was happening, and no one armed him with the tactics that work in such situations.)
The handwriting appeared on the wall when FDA staffers presented what was ostensibly summaries of Swedish Match points and FDA’s assessment of those claims. What they were actually presenting was rebuttal arguments. I don’t think they ever once said “we agree this is true” or “FDA feels like Swedish Match summarized the evidence about this well.” Instead they looked for any excuse to express doubt about the science: cherrypicking outlier studies, subgroup analyses, questioning methodology, asserting that failure to control for some variables means the null results might not be valid (even though controlling for such confounding would inevitably reduce the associations), etc. The presentations were blatantly about creating doubt. (I suspect the top CTP people cringed a little bit when their staff did not do a better job of obfuscating their marching orders.)
Indeed, TPSAC members — even those who would go on to cast negative votes — were voicing a few objections to CTP’s one-sided assessment. “What drives you to look at that data and say the FDA has a concern?” If this is your standard, then “how does anyone ever say…?” “What is FDA’s instruction to TPSAC about what constitutes evidence?” (The answer to the latter was: we are putting that question to you tomorrow. That was actually not true.)
It is hard to imagine that CTP would take this approach if they wanted to keep the door open for approving the application. It will be really difficult for them to come back and say, “TPSAC was hostile toward this application, and our summaries of the evidence focused on why we should doubt all the major conclusions, but we think Swedish Match made their case.”
The other conclusion that was clear from FDA’s rebuttal is that their procedure for company filings is: 1. Give some vague instructions about what they want, and tell the company that the details of how you do this are up to you. 2. Look at the applicant’s attempt to comply with the instruction. 3. Use the details of what they decided to do to form a list of what constitutes inadequate compliance. When Swauger fought back about this, pointing out that there is a disconnect here, with FDA complaining about everything Swedish Match did, even though it was all done to comply with their instructions, CTP (Ashley) replied that we did not tell them to do exactly that research with those methods.
Yeah. Again, this is not about truth-seeking. This is about winning a case when serving not only as both prosecutor and jury, but also the author of the rules of procedure, which you mostly keep secret and rewrite at will.
This is certainly their approach to the Substantial Equivalence process. At the end of the meeting, the top CTP officials all made remarks about how great it was that Swedish Match took this step and how they look forward to more MRTP filings. But the real message was, “you spent ten million dollars in pursuit of some incredibly modest and obviously correct adjustments in a few labels, and look how easily we spanked you for it; anyone else want to try to ask us to do something we don’t want to do?”
The high point of the first day for me was the public comment by Reynolds’s Geoffrey Curtin. He basically fired a torpedo at the application. It is easy to see why Reynolds might have this goal, because their application to make this label change for the entire category was turned down (despite even the likes of McAfee conceding that the evidence does not support those labels) and the Swedish-Match-only change would create a competitive disadvantage. But the tactic was what amused me. Some background is needed to understand that.
Swedish Match presented some calculations from what they call a “model” of behavior (it is really just a very complicated bit of arithmetic). FDA has made noises about requiring population modeling for applications. Reynolds threw a ridiculous amount of money to Environ, a consulting firm, to develop this particular “model”. What Environ came up with (disclosure: in spite of my efforts to try to help them do better) was crap. It was a simplistic calculator, dressed up to look like more than it was with needless cosmetic complications; it made some highly dubious assumptions and spit out meaningless results based on them. The basic tool was not without value, but the authors fiercely resisted making needed changes (e.g., to take just their biggest howler, they assumed, obviously incorrectly and based on no evidence, that smokeless tobacco is 11% as harmful as smoking) — and adding proper epistemic modesty about it — in order to make it valid.
(Aside re peer review: I reviewed the paper in which this “model” was published for two journals. The first time I wrote detailed comments — more than 20 pages — about what needed to be done to make it good. I genuinely wanted it to be as good as possible. I urged the journal to ask for revisions in order to try to increase the value of the paper. But health journal editors do not consider trying to improve the scientific record to be part of their job, and so if they get a lot of comments about how to improve a paper, they just reject it so that they do not have to bother with revisions. So it was submitted to a second journal, with the authors not taking any of my suggestions. I just sent the same review again, with the same advice. The journal again just rejected. The authors presumably wised up and when they sent it to a third journal — the same paper, with no improvements — and asked that I not be allowed to review it, in spite of pretty obviously being the most qualified person to review it. The reviewers that they drew apparently did not understand or care that there was need for improvement, and so it was published in almost exactly the original form that two journals rejected and that an expert peer reviewer called for extensive changes to. Great process, huh? You can understand why I seldom agree to do reviews for journals.)
Anyway, for the last several years, I have repeatedly pointed out the flaws in the Environ model and what needs to be done to create a useful and valid model of consumer behavior. And on the numerous occasions that Curtin was in the audience for that (or was on stage and I was commenting from the floor), he objected and grumbled, and even visibly got angry (which is saying something for someone who is normally extremely nice and diplomatic). It is easy to understand why — his company spent millions of dollars on that load of crap that I could have thrown together in a month. Reynolds did not actually use that work product it for anything as far as I can tell, but Swedish Match threw some money at it and adopted it. And so when Curtin stood up to comment on the application, I rather amused me that he focused his comments on what was wrong with that “model” and what needed to be done to make it really useful. It really amused me that he did so with the exact same points I have been making for the last few years, and that he had not been happy about. It was awesome at so many levels.
Circling back again to the second day, with the context of the first day, we can study the effects of groupthink. The next question to TPSAC was about whether the proposed new warning statement (the one about “no tobacco product is safe but…substantially lower…”) adequately communicates the potential health risks. The discussion centered on trivial issues about the consumer study Swedish Match did, those that were bickered about at length the first day. McAfee brought up a more important point, that the study did not actually measure what the proposed label caused people to believe. Not that any artificial study could ever do that, since beliefs would evolve based on whatever impact the change had on the zeitgeist, and not be merely what someone thought when sitting by themselves and reading the label. So there is no way to meet that standard.
(Anyone who is still reading at this point will appreciate the observation that this kind of thing is what makes McAfee a greater threat to THR, and thus real public health, than the high-profile anti-tobacco extremists who are the lightning rods for social media chatter. The latter are interchangeable clowns, since their playbook consists of making up “facts” (which in this day and age are quickly rebutted) and anyone can do that. But McAfee is smart enough to identify good arguments, whether valid or not. He is particularly good at creating complicated arguments that are false but do not depend on false “facts” — as in the example above — and tricking people into believing that an impossible standard is the reasonable expectation.)
When it came time for TPSAC to explain their votes — all no — it was not the valid point that we could never really measure the effects in advance, but the omission of “WARNING” from the study, that several of them cited. Why? Because the gut feelings evolution has programmed into us say “you can’t lose by just repeating what everyone else seems to agree with — you won’t get fired, you won’t lose friends, you won’t look ridiculous within the group.” Of course, this is a terrible atavistic urge when it comes to doing science, but to fight it you have to be smart enough to recognize the weakness of the groupthink claim and strong enough to speak out hard against it (and you cannot be sitting in a seat where you are not just ignored — Swauger probably would have the chops to pull it off otherwise). It does not require being as good as Feynman, but some tendency in that direction is needed to overcome the bad thinking generated by committees.
The groupthink became even worse thanks to anti-scientific thoughts planted by the supposed consumer representative — who like her predecessor, is a prohibitionist, the diametric opposite of a representative of consumers — Pebbles Fagen. (Yes really. Must…resist…urge. Name jokes are the lowest form of childish humor. Though in this case I think I would have to be forgiven based on the attractive nuisance doctrine.) During the discussion of the label question and the one before, she brought up claims about pregnancy, stating that snus is just as bad as smoking for fetuses. Several member of the committee glommed onto that as their reason for voting that snus is not substantially less harmful than smoking even though (a) the claim is not true in the first place, and (b) even if it were, this would obviously not, by itself, alter the comparison.
This is a perfect illustration of why science-by-committee is such a disaster. One person made one incorrect remark, and the group glommed onto it, or at least refused to dismiss it because that would be politically incorrect. This is in spite of it coming from someone who is not even ostensibly a scientist. (Don’t let that “PhD” after her name fool you — it from a public health school in “health education”. I am not sure what that even is, but I am pretty sure they should not be giving PhDs in it. She has no apparent education or skills in anything scientific. Her undergraduate was in “rhetoric and communications”. Come to think of it, that pretty much answers the question of what “health education” is.)
Empowered by the vulnerability of the committee to her anti-science, she then started claiming that because most of the research subjects were men (because they dominate smokeless tobacco use) and Scandinavia is not very racially diverse, we do not know if the products might be harmful to women and brownish people, and so cannot translate Swedish results to the USA. This was the next question: what evidence would be needed to show that the effects in the USA would be the same as those in Sweden assuming the use behavior was similar. Never mind that (a) human biology, when it comes to oral health, does not seem to vary much between members of the species and (b) anti-scientists can always come up with “but you have not looked at…?” games to pretend we do not know enough (see also: anti-vaxxers, anti-GMO activists, climate change deniers). Claiming that we have not done every possible study is not a scientific argument, but it captivated the groupthink of this “scientific” committee.
By this time, the committee was not even trying to answer the scientific questions. In spite of the bit of the question I italicized, the discussion was focused on questions of whether the behavior would be similar. (The content of that was mostly nonsense.) Fortunately one member, O’Connor, was enough of a scientist to point out that they were not addressing the question. Contrast this with the non-scientist, Fagen, explicitly objecting to the “assumption” that the behavior would be similar. Sadly, most of the ostensible scientists on the committee leaned heavily in that direction too, not being able to perform the fundamental scientific task of asking an if-then question without fixating on their beliefs about the “if” part. (I wonder if they even understand that this what hypothesis testing is, or if they think that is just something that software does?)
I am not even sure what to say about the last hour of the meeting. There were wandering discourses objecting to “use this when you can’t smoke” messaging (which, in addition to being irrelevant to the questions at hand, is forced upon smoke-free product manufacturers because they cannot make truthful claims about the products being low-risk). There was a vote on whether the evidence supports whether US “tobacco users” (no, it did not say “smokers”) would will switch if the changes were made. None the six who said the evidence does not support that noted the real reasons for that — that the change being proposed is far to weak to encourage THR. Indeed, most of the reasons given were non sequiturs. O’Connor, the only dissent, once again seemed to be trying to answer the question that should have been asked rather than the one that was, pointing out that the data from Sweden suggests that people will switch if they know the truth.
The next question was worse. The following one was literally nonsensical. I will spare you the details. What stuck me the most at this point is that the committee bickered about the wording of good and valid questions but voiced no objections to terrible questions (again, with the exception of O’Connor, but he did not press the issue).
FDA then asked them if the lower risk should be communicated in the warning label, or if it would be better offered as a separate statement. That is a valid question, but kind of strange given FDA’s repeated message that the decision had to be about the application as made, and such a proposal was not part of the application. McAfee suggested it would have been better if Swedish Match had applied for permission to full-on market the product as a low-risk alternative, calling the request at hand a “tempest in a teapot”, implicitly conceding that no one really cares what the warning labels say. That is a valid point, though if you cannot even get TPSAC and FDA to agree that the warnings are wrong, what chance is there of that?
There were four final bits of black comedy. The penultimate question was about recommendations for conducting postmarket surveillance. Nothing useful came of it, but it was amusing to listen to them talking about use social media as if they were visitors from another planet who had heard that there is the thing called social media that seems to be a big deal. The e-cigarette astroturfing claim came up there, because these people cannot stay on topic to save their lives.
That illustrated why the ultimate question that followed was funny: FDA asked if there were any other issues that were not covered in the previous questions. This might have been worth asking if the committee had the discipline to stick to the questions before that.
Then Tomar voiced the observation that perhaps FDA should set up better guidance for what standards the applicants should meet in their applications. Apparently he did not get the memo that suckering applicants down a path and then declaring it to be the wrong path was considered a feature, not a bug.
And finally everyone patted themselves on the back for moving the MRTP process forward and thanked Swedish Match for their contribution to that. I assume they all had a good laugh about that once they were off camera.