Complaints about conflict of interest are designed to maximize conflict of interest

by Carl V Phillips

A colleague who found himself the target of the “you have a conflict of interest!” bullshit game was invited to write his analysis of the nature of conflict of interest as it relates to e-cigarette research. He asked me for input and several points occurred to me. I am writing them up here, in an admittedly disorganized fashion, for possible use in that project.

My first observation is that implicitly endorsing the FCTC-style rules (“thou shalt not even so much as talk to a tobacco company”) but demanding you be recognized as slipping through a loophole (“but I only took money from e-cigarette companies that do not make any traditional tobacco products, so that does not count!”) is a sucker’s game. Similarly arguing “it was only a little bit” or “it is all in the past” is a sucker’s game. It is either foolish tactics or Stockholm Syndrome. Those approaches implicitly concede that the fundamental basis for the ad hominem abuse is legitimate, and then beg one’s attackers to forgive a particular transgression because of a technicality.

But why would they ever want to do that?

Their game here is not based on any legitimate consideration that leaves room for saying “that other research/funding/consorting is bad because of X, but mine is fine because of Y.” There is no X. Research conducted by the tobacco companies and those they fund is among the best and most impeccable in the space. Indeed, I would argue it is the best, but even if you do not go that far it is easy to observe that none of that research (as far as I know) for the last several decades has been shown to be fundamentally deceptive in its methods, analysis, or conclusions, or even just slipshod, whereas there is relatively little coming out of tobacco control that does not fit some or all of those descriptions. (Purely indy and e-cigarette industry funded stuff tends to fall in between.)

Yes, the ANTZ rationalize their claims based on perversions of science committed by the industry, but those took place two generations ago. Such tenuous rationalization is enough to convince their useful idiots, but it is clearly not the genuine motivation. After all, Volkswagen was part of the Nazi establishment four generations ago, but that does not mean that today… ok, bad example. Actually, what that example illustrates is that compared to many other industries, the current tobacco industry is remarkably free from active deception and calculated malfeasance. (That is not to say that they do not use their political influence to further some aims that are not in the public interest, as every powerful industry does, but that is quite distinct from whether the science they help produce and the scientific claims they make are legitimate.)

So what is the ANTZ’s real motivation? It is to shut out the voices of anyone who disagrees with their special-interest political beliefs. It is also to minimize the scientific scrutiny their junk science is subject to. And it works. Even though the industry conducts and sponsors some of the best research on the topic, they very seldom attempt to rebut the tobacco control junk in the science space (they do in the litigation space, but that is a different story and arguments there are supposed to be as biased as possible). Most of them are scared to openly disagree with anyone. This was not always so, but ~17 years ago, they abdicated their responsibility of policing the junk science that was arrayed against them. The result, in terms of escalating junkiness, is just what you might expect.

Personally, I think there should be a rule that you can only offer analysis of the comparative risk of various tobacco products or of how to assess whether there is a gateway effect if your name is “Carl”. How could that possibly be justified? Pretty much the same way the ANTZ justify their censorship, only it is true in this case: There has been an enormous amount of blatant garbage written on those topics by people not named “Carl”. Most of them have patently misrepresented their results, one way or another, and hid pertinent information from the reader in order to further their agendas. Clearly they cannot be trusted. If I could get that rule adopted, it would make my life very easy. I could shovel out whatever garbage I wanted on those topics, and no one would be allowed to challenge me. (Of course my readers might surmise that I probably wouldn’t pursue my self-interest and take advantage of that opportunity, but the point is that I could.)

So, bringing this back in, those that have cynically and immorally imposed this quasi-censorship would only change their tune about purely e-cigarette companies if they did not want to censor them too. But they do, of course. They might have a slightly more difficult time selling that censorship to their useful idiots based on the rationalization of ancient malfeasance by entirely different companies, but only slightly — they really are idiots for the most part, after all, and it is not as if today’s cigarette companies are the same entities they were in 1970. So begging for an exception, as if the censorship and ad hominem were reasonable in other cases, gives away all the space available to fight this absurdity.

So where to fight? A good starting point is to note that the main effect of this tactic is to ensure that almost all research suffers from the same strong conflict of interest (COI) that results from being a member in good standing of the tobacco control industry. It is hard to do any real research without funding, and basically the only source of funding are (a) industry and (b) tax money that is controlled, one way or another, by the ANTZ. There is no “neutral” funding in this space. Notice that researchers who try to work in this space without depending on one or the other of those mostly (not to say always) produce mere commentaries or simplistic studies.

Keep in mind that the most common COIs in any politicized science arena stem not from financing, but personal beliefs and goals. (And what exactly is a COI? I have written about that before (e.g., this paper, this blog post), and get back to it below. I told you this was disorganized.) Whenever you see someone making a big deal about never having taken money from industry, your first thought should be “so what — what really matters is what you want the world to look like.” (My other thought is always that such protests reminds me of someone who tells you he has not touched a drink in eight years: He is not telling you something about the nature of alcohol, but about himself. He is saying that, unlike most people who can drink with no problem [have enough of a moral compass that they do not just sell out to whoever pays them], he fears that if he personally has one drink [takes funding] he will melt down.)

Everyone who is part of the tobacco control industry has demonstrated a particular political bias. If the ANTZ can keep everyone else from having the resources to work in the space, then everything that is produced comes from the perspective of their interests. Having people with opposing political biases all doing science in a space is a poor substitute for the ideal of only having people who diligently pursue good research and analysis, wherever it takes them (there just are not very many of us :-). But the presence of opposing COIs at least beats the heck out of just having one faction.

This is not to say that the ANTZ COIs are all non-financial. Quite the contrary. Much of the discussion on this topic is motivated by people trying to make sure that any association with industry does not destroy their access to the tobacco control funding spigot. So the present discussion tends toward questions like, “As long as I just work with the pure e-cigarette industry, I can still count continuing access to that gravy train, right?” (Ans: Nope) or “How about if I avoid any positive engagement with industry, but still say something good about THR?” (Ans: Maaaaybe, so long as you do not say anything good about snus and insist that e-cigarettes have to be heavily regulated, to the point that they are effectively conceptualized as medicine). It is difficult to imagine a greater financial conflict of interest than the threat of losing all future funding if you deviate from the party line.

Difficult to imagine for indy researchers, that is. Because employees obviously have a similar COI, and that includes not just industry employees, but the many employees of ANTZ operations like CDC, FDA CTP, NIH, WHO, ACS, BMA, ET AL. People working on anti-THR in those operations outnumber and vastly outspend all the genuinely independent researchers who merely have the COI that results from chasing grant money. These organizations have clear institutional anti-THR policies that their employees would dare not violate.

In short, these attacks on COI by the tobacco control special interest are designed to make sure almost all the COI points in the direction of their favored policies. And it comes very close to succeeding.

It is also important to note that funding or beliefs of any sort can only create a COI on a case-by-case basis. For there to be a COI, there needs to be a conflict. Wait, you might ask, between what? After all, if you just listen to the ANTZ talk about it, you might think COI was an existential personal characteristic of the individual, a state of being a sinner akin to how some people classify being gay: If you are such a sinner, the only acceptable behavior is to be celibate/silent. They define COI — by which they mean being a person who ever had any constructive engagement with industry (unless you are so thoroughly a member of their club that you have earned a bit of immunity) — such that it exists if what you are writing is a hotel review on TripAdvisor. Obviously, for that particular report, one’s opinions about tobacco products are not a conflict. But, again, conflict with what? In this case, with the implicit promise that you are telling the truth, as best you can, about that hotel. Obviously there is not a state of being a sinner that creates a COI there.

What those political or financial interests — on either side — might conflict with is, for example, the implicit promise one is providing an honest assessment of how hazardous e-cigarettes seem to be. Someone intent on getting $3.5 million more for Portland State University from the anti-tobacco spigot might cook(!) their research to claim a non-existent risk; someone who wants to sell their e-cigarette products might decide downplay uncertain possible hazards; and someone trying to thread the needle and portray e-cigarettes as too harmful for unfettered casual use but still great as a cure for smoking might decide to declare that they are 5% as harmful as smoking. In all of those cases, the suggested financial or political goals of the authors conflict with the implicit claim that they are offering the best scientific estimate. (In none of those cases, it should be noted, can we conclude that the hypothetical conflicting motive is cause of whatever was written. Just because an incentive exists does not mean that a reported claim is not a genuine attempt at honest science.) The rare exception on this point would be — ahem — a consumer advocate who wants to give consumers the most accurate possible information to make informed choices; being a consumer advocate is a political interest too, but it is not a conflict of interest in this case because it happens to align perfectly with the goal of providing the best scientific estimate.

By contrast, an advocate for e-cigarette consumers might have a conflict when addressing the question of whether there is a gateway effect. If he did not consider the possible gateway cases to be among those consumers he seeks to benefit, then he might downplay that risk. Similarly, if the Portland State ANTZ (PortlANTZ?) claimed not that they were assessing the real risk from vaping, as they did, but rather clearly stated that they were setting out to show whether an e-cigarette can be forced to operate in a way that produces a lot of formaldehyde (which is what they were really doing), then there would not have been a COI. They would have been stating that their mission was to try to make e-cigarettes look bad, and they would have been doing just that; the fact that they stood to rake in a fortune for doing so would have been motive for doing such research, but not a conflict with it.

The general point here is that having a COI can only refer to a particular act (a particular bit of authorship). It does not exist as a free-standing characteristic of individuals. Any legitimate claim of COI needs to be phrased as (or at least can be rephrased as) “you have incentive X to deviate from your [implicit] claim that you are doing Y because of Z” not just the equivalent of “you have cooties.”

Finally, it is worth pointing out that the past cannot create a COI, only the future. That is, having funding from industry or from NIH is not a conflict. Nor is having had it in the past. That deal is done. It creates no further incentives unless it came with a promise of producing a particular result. It is wanting more funding from them, or a job from them in the future, or wanting to keep your present job, that creates incentives that might conflict with doing honest science. It is worth adding that the industry would almost certainly not cut off future funding to someone who said something “wrong”, but the ANTZ would and do, and that industry grants tend to be as unrestricted as possible, while ANTZ funding agreements often come with a promise of what the results will show.

So what do we do about the problems created by this pervasive COI in our field? That’s easy: Insist that everyone with a COI — and we can just round that up to everyone when we are talking about a topic like this, to avoid the need to develop a Platonic scientist detector — provide enough detail about the basis for their analysis that we are not just taking their word for most of it. If no one is to be just trusted, then everyone should provide an audit trail — report their exact methods, make their data available, show their work, do sensitivity analysis, stuff like that. The funny thing is, though you would never know it from reading health science, that this is exactly what scientific researchers are supposed to do anyway.