Tag Archives: “safe”

TrANTZlating “no safe level”

posted by Carl V Phillips

I was asked by a CASAA member to turn the trANTZlator on the very confusing phrase, “there is no safe level of X”.  You have probably heard or read that phrase most often in the context of “there is no safe level of second-hand smoke.”  This is actually a pro-THR lie, as mentioned yesterday but the same phrasing is frequently used as an anti-THR lie in the context of the chemical exposures from THR products.

The reason this is a lie is that most anyone reading this phrase interprets it in the natural way:  no matter how low the quantity of the exposure is — and in particular, including whatever level of exposure is being discussed in the particular context — there are health effects that are significant enough to warrant being called “unsafe”.   But this is never what the phrase actually means.  It could be argued that the phrase is fine to use in a scientific context because readers there know what it really means.  But it clearly should never be used in popular communication where it will almost always be misinterpreted.

Those who use it in popular contexts either know that it is going to be misinterpreted (and thus are intentionally misleading people) or actually think that the sloppy technical phrase has its natural language meaning (in which case they have no business claiming they know what they are talking about).  Most readers of health science figure out the truism that quantities matter or “the dose makes the poison” — i.e., that trivial exposures do not matter in any practical sense.  The “no safe level” claim sends the message that this (true) belief is wrong, and it is difficult for the lay reader to have the confidence to reply, “I know that is never true.”

Most of the time, the claims “no safe level” is a sloppy shorthand (understood in the scientific context, but unknown to the broader audience that hears the phrase) for “no established safe level” or “no known safe level”, which in turn trANTZlates into “there is almost certainly a level of exposure that is perfectly safe, and it may well be that the levels being discussed in the particular context would be considered safe (in the common sense of the term), but the science has not been done.”  In other words, what is really a statement of ignorance is intentionally presented in a way that will be interpreted as a very strong affirmative claim.

Very occasionally the phrase actually does mean “we have tested exposures as small as we possibly could, and we still saw an effect” (example).  But even then it does not have the meaning that most people take from the phrase.  “As small as we could test” is still not “as small as could occur”, and “an effect” (e.g., at the molecular level) is not the same as “a measurable health effect” (i.e., real results on the actual health of people).  The former of these puts the lie to the literal interpretation of the statement, but it is the latter that really matters.  Any effects that are measured down to the lowest practical levels of exposure are not health effects, and only actual health effects can warrant use of the word “safe”.  Tiny measurable effects in vitro or even at the biomarker level do not necessarily imply actual health effects, and if they do, those health effects are almost certainly undetectable because they are tiny or extremely rare.

That leads to the ultimate problem with the phrase.  The natural interpretation of something being unsafe (which it must be, if there is “no safe”) includes the harm/risk having nontrivial magnitude.  What probably deserves to be called the #1 anti-THR lie is the statement “not a safe alternative to smoking”.  It literally means “there is some tiny tiny risk” or even “we are not sure it causes absolutely no risk whatsoever”, which is true for every exposure and thus technically true, but obviously misleading (i.e., the worst kind of lie).  When people read “not safe” or “no safe level” they do not interpret it as “some utterly trivial health risk that is so close to zero it does not matter”.

Indeed, there are a few exposures where any exposure actually does create a health risk because the risk is a constant role of the dice.  The simplest example is exposure to motorized transport:  Each second of exposure causes a tiny finite risk of injury from a collision.  The degree of safety is proportional to the quantity of the exposure, with any positive amount of exposure creating a positive amount of risk.  But if someone said “it is not safe to drive or walk to work”, it would be immediately clear that they were not using those words in the normal way (and so if they said the exact same thing in a context where people would believe them, it would be a lie).  If they said “there is no safe level of driving”, they would be effectively communicating that the phrase does not mean what people think it means.

As a slight aside, this reminds me of an ironic story that I recently heard (from a friend who contributes to this blog) about people who are so afraid of the “no safe level” of ETS will cross the street twice to avoid catching a whiff from a smoking area.  But we calculated that the risk from the average street crossing is three, maybe four, orders of magnitude higher than the risk from that ETS exposure — even if the effects of exposure are linear.  But furthermore, unlike the risk of accident from crossing the road which is pretty close to linear over the amount of exposure (crossing the road once is 1/1000th as risk as crossing 1000 times), chemical exposures like ETS tend to be far less than linear (that is, even if an exposure has risk X, an exposure that is 1/1000 as great has a risk much less than X/1000, and perhaps basically zero).  Of course, the total health damage done by the “no safe level” message in that case is pretty trivial compared to the damage done by using the phrase to discourage THR.

So what does “no safe level” mean?  Unfortunately, the trANTZlator can offer no consistent substitute.  The phrase is sloppy and can mean any of several things, ranging from “we are pretty sure there is a safe level but are not sure what it is” to “the risk is proportional to the exposure level and it never goes to zero, but it gets as small as you want, and well into the range that people would consider safe”  But it never means what most people think it means, that the quantities being discussed are known to cause health risk at a level that we should care about it (and moreover that any exposure causes such a level of risk).  After all, if the liars thought they could defend that claim, they would simply make that claim.