Chris Snowdon recently posted an excellent analysis of a recent New England Journal of Medicine article that calls for bans on smoking in non-single-family homes. This has all sorts of interesting implications including class-based discrimination (I would bet that most of the wealthy activists who are calling for these restrictions own detached houses and so would be immune from such restrictions they want to impose on others), ongoing nonsense about the effects of ETS, and various other matters. But I will highlight just two points.
The penultimate point is what this says about the quality of health science journals. As I noted a couple of years ago anti-tobacco activism is a threat to the already rather shaky integrity of health science and publication thereof. In this case, Snowdon discovered that the references for the key empirical claims in the article were to one document and one speech which did not actually support the claim. (His impressive abilities in forensic research were probably not too taxed by this investigation.) Many people believe that “prestigious” journals in health science afford some measure of scientific legitimacy, though I have noted before that when I want a teaching example of a dumb excuse for epidemiologic analysis I usually start with NEJM. (This is probably due to this and other medical journals depending on editors and reviewers who were trained as physicians — consumers of research, never trained as scientists themselves — and do not understand that they do not understand scientific inference, but that is a story for another time.)
Nevertheless, it is important to note the implications of yet another case of anti-tobacco extremists taking advantage of the inability of editors and readers of health science journals to recognize complete anti-scientific nonsense when it comes across their desk. By doing so, they undermine the trust that is the key to science.
That was the second-most important point. The most important is that Snowdon is now a solid front runner in our contest to predict the most ludicrous anti-tobacco junk science article that will appear in the latter part of this year. He predicted that there would be a study that claimed ETS travels through phone lines, and notes in his post that the article in question claims that ETS travels along electrical lines, which is tantalizingly close. We have to give him the inside track right now, but it should be noted that the NEJM authors were actually probably claiming something different: The claim is probably that tiny a bit of ETS can get into an electrical box (if it is not well covered), and from there an even tinier bit can get through small holes in the box (or, if the building is a substandard firetrap, there might be no box, but then — as Snowdon points out — the residents have bigger problems), from there a tiny bit is in the conduit or dead space in the wall, and from there an even tinier bit can get through small holes into the neighbor’s box, and then….
Come to think of it, maybe they were claiming that it transmits down the wires — it seems just a plausible. Full marks for Snowdon then. But the contest is still open for entries (though you do not get credit for anything that came out before you “predicted” it, of course) and who can doubt that something even more outlandish might still get written — the extremists’ creativity in making claims that damage the legitimacy of health science knows little bounds.
– Carl V. Phillips