How to communicate the implications of negative evidence

by Carl V Phillips

Continuing the series I started yesterday on how to mobilize the truth for THR advocacy.  I believe the next logical point for the typical discussion with an open-minded but uninformed THR skeptic, after yesterday’s, is the inevitable discussion about “addiction”.  However, I will put that off and go over some thoughts I have been assembling for an interview this afternoon.  The planned topic is what sources vaper advocates can use when arguing the merits of THR to MPs (it is a UK show).

The discussion will probably also address the topic of what arguments to use also, of course.  But what about those sources?  What form does our evidence take?

I think there are three categories of scientific points to be made: 1. how do we know it is low risk? 2. how do we know it works for smoking cessation? and 3. a smattering of others.

Our knowledge about the low risk of long-term smoke-free nicotine use comes primarily from studies of Western (Swedish and American) style smokeless tobacco.  Those products deliver nicotine (and other exposures too) and have been definitively shown to be very low risk.  There is a substantial epidemiology on this and a few good summaries of parts of it.  Due to the limits of the science, we cannot say for sure whether ST is 1% as harmful as smoking, or 2% or 0.1% or even 0%, or even a little bit good for you.  All of those are plausible given the evidence.  However, a number as high as 5% is not plausible.

We can extrapolate our knowledge about ST to pharmaceutical nicotine products (“NRT”), which are quite similar to e-cigarettes.  We have to extrapolate because we do not have the same evidence about long-term pharma nicotine use (because the regulators and manufacturers have pretended there is no long-term use, they have avoided studying it).  Everyone other than anti-nicotine zealots is comfortable in concluding that pharma nicotine has the same low risk as ST — perhaps 2% as bad as smoking, perhaps 0%, but almost certainly not far out of that range.  We are reassured in this extrapolation of the data by the fact that there are no apparent serious acute effects from pharma nicotine (of reasonable doses), nor has there been a strange pattern of disease detected among the many long-term users.

We can then apply this same extrapolations to another way to deliver pharmaceutical nicotine, e-cigarettes.  They are, of course, little different from the pharmaceutical nicotine inhaler which we are confident is low-risk and which the UK’s MHRA has approved for long-term use for THR.  As with pharma products, we have enough experience with the use of current-technology e-cigarettes to be confident there is not a substantial risk of acute effects and there is nothing unexpected that seems to appear after a few years.  E-cigarettes are enough different from smokeless tobacco that we cannot be sure they are quite as low risk.  They might also be even lower risk, though not much lower obviously, since there is not room for much lower.  I think the smart bet would be that they are a bit higher risk (due to airway involvement, primarily), but you would probably never be able to collect on that bet because it is too close to call.

So, back to the issue of what to say.  Any scientist (a category that does not include most people in “public health” who pretended to be scientists) will recognize this pattern of evidence and be able to use it:  Almost every conclusion in science is based on this analysis:  What we have observed in a well-studied similar case suggests that D will not be caused by E (where D=major health risk; E=e-cigarette use).  We have observed E itself to a limited extent and have seen no reason to doubt the extrapolation.  Therefore it is reasonable to conclude that E does not cause D.  Exactly how confident to be about this depends on the similarity of what we have studied, how much evidence there is, etc.  Science, it turns out, is more art than science in situations like this.

How can we explain reasoning like this to the average interested person?  I am not even sure I succeeded in explaining it to you, my readers.  I see no choice but to appeal to the conclusions that can be drawn from the lack of evidence, but we can keep it simple:  “There is overwhelming evidence that smoke-free alternative nicotine products pose about 1/100th the risk from smoking.  There is no reason, based on either our substantial knowledge about what vapers are exposed to, or actual observed health outcomes, to doubt that this applies to e-cigarettes.”

At that point, you should add in the point I made yesterday.

So how to respond to the inevitable reply of “but there is no proof they are safe!”  First, of course, is to point out that no one is claiming they are 100% safe, just that they are about as low risk as drinking soda or eating french fries (sorry, “fizzy drinks” and “chips”), or commuting to work, and a miniscule fraction of the risk from smoking.  But do not stop there.  The next bit is tricky but crucial.  Most people involved in this debate are not scientists, but it is necessary to understand one bit of science:  All statements that something has low risk are extrapolated from other observations about something that is similar but not exactly the same.

When a product has been specifically studied for a long time (like snus has), the extrapolation can be made with a bit more confidence.  But it is still an extrapolation.  Snus has been shown to be low risk when used in the past (in particular forms, in particular ways, by particular populations), but that cannot prove that snus use in the future will not be high risk.  We can be pretty confident but there is no proof.  Similarly there is no proof that eating apples will not, in the future, turn out to be high risk.  And but the same token, we cannot be sure that smoking will not be lower risk in the future than it has been in the past (indeed, it almost certainly will, due to improved disease treatment).

So every conclusion about risk is an extrapolation from the most similar experiences that we have actually studied.  The extrapolation from other sources of smoke-free nicotine to e-cigarettes is very convincing to anyone who thinks like a scientist, but it cannot be proof.  So, how do you offer a simple citation to support this?  You cannot really.  Sorry.  It is necessary to appeal to the totality of what we know about the topic, including what has not been published (i.e., no discoveries of a reason to suspect that e-cigarettes might be higher risk than expected).

I will post this and try to address point 2 in another post before the show today.  The good news is that point 2 is one that can be demonstrated based on evidence rather than its lack.

2 responses to “How to communicate the implications of negative evidence

  1. Actually, I’d say drinking fizzy drinks would stand a fairly strong probability of being higher risk than vaping e-cigs. We *know* the harm from drinking sodas, and we know a moderate amount about the mechanisms of that harm (at least for sugar drinks — the harms from the sugar-free ones aren’t as clear, but I have a feeling they’re more substantial than recognized.) We don’t know of any harm at all from e-cigs, despite “field epidemiological trials” (of a sort) involving literally millions of subjects through billions of exposures. We also don’t see any substantial arguments out there for theorizing harmful mechanisms that would be likely to cause problems in the future.

    – MJM

  2. Pingback: The value of individual testimonials and more on negative evidence | Anti-THR Lie of the Day

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s