Asking the right questions (retitled)

I’ve been criticized in the past for focusing on criticism of bad health reporting, rather than aiming some positive reinforcement at the good pieces.  Well, here you go.

The story (byline Genevra Pittman) is about lasers used in smoking cessation, an idea with limited plausibility (unless…you suspend tobacco execs over a tank full of sharks with frickin’ laser beams attached to their heads, or something).  The idea isn’t new (lasers, not shark tanks); ads for laser smoking cessation have been in papers and the internet for years.  Tobacco is a tenacious and deadly addiction, and treatment of this addiction is often unsatisfying.  The use of lasers and other “acupuncture-like techniques” to assist in smoking cessation is popular and profitable. “But does it work?” asks the Reuters piece.

There are a couple of ways to answer this question: a flood of anecdotes, or a systematic evaluation of the plausibility of the claims, and of the data.  Advertising normally focuses on anecdotes.  This approach has the advantage of allowing people to pick the stories they like, highlight them, and use them to promote their business.  Data are much trickier, and more fickle.  They can lead you, as a promoter of lasers, to places you may not want to go.

Does it work?

The Reuters piece gives a good summary of the claims made by a laser smoking cessation clinic.  Laser smoking cessation is basically a variant of acupuncture. Rather than needles, low energy lasers (low enough energy to have no sensible effect on tissue) are aimed at certain “points” believed by practitioners to have some sort of physiologic significance.  The article cites these claims in an objective, rather than promotional tone, then asks the Right Question.  Does it work?

That question is deceptively simple, and often asked and answered improperly in mainstream media pieces. Pittman examines the source cited by the laser promoter, summarizes its findings, but rather than stopping there, she asks the right expert.  She found one of the authors of the Cochrane Collaboration’s systematic evaluation of acupuncture techniques for smoking cessation.  She explains his caveats about the single citation supporting laser therapy and problems with the study, including its reliance on patient self-reports.  But one question was left out of both her report, and that of the Cochrane Collaboration.

Can it work?

Cochrane came to the (likely correct) conclusion that the use of acupuncture (including laser therapy) has not been shown to aid in smoking cessation.   The Cochrane Collaboration has been an invaluable resource for evidence-based medicine, but, like most EBM resources, often fails to ask one important question: can this even work?

This is what many of us see as a fundamental flaw in the current approach to evidence-based medicine.  EBM certainly understands the concept of probability, but often fails to use it properly.  While EBM will talk about Bayesian statistics and likelihood ratios, it often fails to take into account what they mean when it comes to implausible claims.

Implausible claims are those that, if confirmed, would cause us to re-examine our fundamental understanding of chemistry, physics, and biology.  The classic example is homeopathy, an 18th century belief system that has become popular among many alternative medicine practitioners.  It relies on a form of vitalism, a made up “Law” of Similars, and if it truly worked, would require us to abandon proved concepts such as the law of mass action (and our understanding of physics as a whole).

The problem with implausible claims (aside from their implausibility) is that when tested, they can, through chance alone, lead to positive results.  This allows supporters to cherry pick favorable data, despite the fact that these data are likely false positives.

If we were to create a hierarchy of implausibility, homeopathy would certainly be near the top.  Acupuncture would be just a bit lower.  It’s not inconceivable that sticking needles into someone (or zapping them with lasers) would have some sort of physiologic effect.  According to the owner of the laser clinic, it works like this:

“When you smoke a cigarette, you artificially tell your brain to release endorphins,” Frank Pinto, the owner of Innovative Laser Therapy, told Reuters Health. Therefore, quitting leads to a quick drop in endorphin levels, he said.

“The laser basically stimulates the nerve endings to tell the brain to release a flood of endorphins” to boost a patient over that initial 3-5 day hump of withdrawal symptoms, he said.

That sounds superficially plausible (one study in rodents found the neurotransmitter adensoine released after the animals were stuck with needles). But if it’s true, then any noxious stimulus could have the same effect, rendering the need for identifying non-existent “meridians” and the use of  expensive equipment moot.  But laser therapy seems to be even less plausible than other acupuncture modalities.  According to Cochrane:

Low level laser therapy produces nosensation, and there is still some uncertainty whether it has a physiological effect on normal tissue. From the researcher’s point ofview, laser therapy has the advantage that both patients and practitioners can remain masked to group allocation by using defunctioned laser apparatus.

In other words, placebo-controlled studies are easy because you can’t actually tell if the laser is on or off unless you peek at the switch.

The Reuters article by Pittman is a fine example of how to approach unusual or hyperbolic health claims.  Its primary deficiency is one shared by many scientists examining such claims: a failure to ask “is this even possible”.

5 Comments

  1. If you look at the study, more people in the control group (8) reported having ceased smoking at 18 months than at 6 months (5) and 3 months (5). This is curious since for the 18 month follow up on all 340 participants only 40 replied.

  2. FWIW, I would have put traditional acupuncture quite a bit lower on the implausibility scale. Of course it’s a moot point, because there’s plenty enough data now to consign it to the dustbin of failed ideas… but without any data, I wouldn’t put the a priori probability anywhere near as low as homeopathy. Homeopathy would require that vast swaths of physics and biology were dead wrong. Acupuncture would not require any of that; it would only require that there was quite a bit of biology that is still unknown (likely true) and that somehow the ancient Chinese had figured out some of what we don’t know (not so likely, but who knows)

    Laser acupuncture would be right up there with homeopathy, though. At least when you stick a needle in someone ,we know that produces a definite physiological response, albeit a trivial and probably useless one. At least there is a genuine input to the system, so the idea of it producing a certain output is not entirely unbelievable. With a low-power laser though… yeah… that’s just crazy.

  3. Maarten Bergvelt

     /  September 30, 2010

    You say: “vitalism, a made up “Law” of Similars,”

    All laws are made up. Or do you have another source for laws of nature, than people making stuff up?

  4. Are you being pedantic, or do you really not understand? I need to know whether to answer with patience or scorn.

  1. Quick Links | A Blog Around The Clock
%d bloggers like this: