FDA does not understand how science and knowing work

posted by Carl V Phillips

Epistemology (roughly, how we know what we know) is at the core of scientific inquiry.  Some of the most profoundly anti-scientific claims that you will ever see — certainly among the most frustrating — are attempts to replace the complicated epistemic processes in science with simplistic rules of evidence.  Typically these are so oversimplified as to be suitable only for a stylized game like high school debate contests.  The rules are too simplified even for a courtroom, the context for which they are often designed.

As I noted in the first of this series of posts on the US FDA, FDA’s epistemic toolbox pretty much contains only one tool: their standard recipe for doing clinical trials of new pharmaceuticals.  That works pretty well for what it does, but it obviously represents only one little tiny corner of scientific methodology, let alone scientific knowledge.  The FDA document that I started discussing yesterday makes clear how unsuitable FDA’s limited understanding of science is when it comes to studying THR.  It includes:

Because clinical studies about the safety and efficacy of these products [e-cigarettes] have not been submitted to FDA, consumers currently have no way of knowing

This is followed by some bullet points, which I will address shortly.  But for now, just consider that sentence fragment.  No way of knowing?  Are they kidding? There is no way to complete that sentence truthfully, other than with something circular like “…no way of knowing what clinical studies submitted to FDA do say.”

The mind boggles at the arrogance.  The typical epistemic error is to claim that there is no evidence in the absence of studies that follow a particular clinical study recipe.  That is also completely wrong.  But there is something extra special about a unit of the government claiming that knowledge does not exist until it is formally submitted to them.  That is downright Soviet, or maybe even more l’état, c’est moi.

Since I cannot think of much to say about that other than shaking my head in disbelief, I am just going to pretend they said, “Because clinical studies have not been done….” and explain why this — the all-too common misconception that clinical trials are the best way to answer every single scientific question — is a lie.

The bullet points that complete the sentence are:

• whether e-cigarettes are safe for their intended use,
• what types or concentrations of potentially harmful chemicals are found in these products, or
• how much nicotine they are inhaling when they use these products.

For some questions it takes a bit of work to explain why clinical studies are not the most useful method for finding an answer.  For example, clinical trials are not very good at answering the most important questions about whether smokers will switch to a THR alternative.  But that takes a bit of explaining (which I am sure I will eventually cover in this series).  But for now, FDA has provided examples that are so obviously wrong that there is no such challenge.

How do we know how much of what chemicals (nicotine and others) are in the vapor?  A clinical study is not well suited to answering this at all.  Instead, we need a method for drawing out and analyzing vapor that avoids the complication of the user absorbing some unknown portion of the chemicals (i.e., which does not include a clinical component).  And that has, of course, been done.  FDA even did it themselves (though they lied about the results).

We could replace the last point with something that actually does involve the user, like how much nicotine users are taking up?  That is possible to answer using clinical research methods.  Possible, but not necessary.  One way to measure nicotine uptake is blood chemistry.  But another is to just be the person taking up the nicotine and feeling the effects.  That will not answer some questions very well, but it is clearly a way of knowing.

How about whether e-cigarettes are safe (I am not sure what the “for their intended use” clause even means)?  Well, in that case some clinical studies would have been useful and would still be useful.  Indeed, we would have those if those in government “public health” thought of themselves as public servants (who would observe that the public wants this information and then try to provide it in order to possibly improve health) rather than public masters.  Nevertheless, clinical studies are obviously not the only way of knowing.

We have learned an enormous amount — in spite of governments’ uselessness in this matter — about the health effects of e-cigarettes.  We have ample data to rule out important negative acute effects, to be confident there are no non-rare major short-term health risks, to be optimistic about the lack of long-run effects, and to be quite sure that they are better than smoking by a lot.  This evidence comes from the experience of users as well as extensive data about other smoke-free nicotine sources and our knowledge from various sources about exposure to the other chemicals.

The general lesson here is that when someone claims “we know nothing” or “we have no evidence” about some scientific question, they either have no idea how scientific inquiry really works or they are lying.  Or both.