Articles

Science-Based Medicine 101: Plausibility

In part 2 of the Science-Based Medicine 101 series we take a look at the second pillar of good science: plausibility. This blog post was written for a lay audience so more advanced readers will need to indulge me here…

***

I really enjoy sci-fi action movies. I love the convincing special effects and the fact that heroes can accomplish the physically impossible without skipping a beat. Implausible events unfurl with convincing reality, and you never know what might happen with the plot.

I also enjoy the TV show, America’s Funniest Home Videos, for different reasons. The mundane nature of actual reality, and the often predictable, but hilarious mistakes made by those I relate to result in some pretty hearty laughs.

But there is a big difference between these two forms of entertainment: science-fiction requires the suspension of belief in plausibility, while home videos are based on plausible outcomes. When it comes to medical research, though, plausibility can mean the difference between science fiction and reality.

Over the centuries, scientists have eagerly concocted explanations for medical phenomena. However, it wasn’t until quite recently that our knowledge of science has expanded into an amazingly concrete and fairly deep factual account of how things work. I credit the scientific method for these astonishing advances that have vastly improved our quality of life. Without rigorous testing, repeating, and retesting in a controlled manner, we wouldn’t have many of the life-saving medicines, procedures, devices, and treatment options that we have today.

And yet, we really like science fiction. We like to believe in magical solutions and discoveries that break the mold – that there could be something out there that we find that no one else has thought of yet, or observed, or harnessed. This drive to discover is a wonderful force for investigation and scientific advancement, but it is a double edged sword. The other side can result in irrational beliefs, magical thinking, and snake oil science.

Take for example the field of homeopathy. Founded by German physician, Samuel Hahnemann in the 18th century, this system of medical treatment is founded upon the principle that like treats like, and that a medicine is at its strongest when it’s at its weakest possible concentration. So, for example, if you have cold symptoms (runny eyes and nose) the treatment would be to drink onion juice (because onion fumes produce runny eyes and a runny nose), but at a concentration so dilute that there is actually not even one molecule of onion extract left. Hahnemann believed that water has memory – and that the fact that water molecules had been recently in the presence of onion molecules gave the water healing powers. In fact, the act of shaking the water made it even more potent as a therapy.

Now, I have no doubt in the sincerity of Hahnemann’s beliefs. I also have no doubt that homeopathy (for those who believe in it) can have a strong placebo effect. But to believe that there is an actual physiological effect of shaken water for the treatment of all sorts of diseases, is to deny everything that is known about human physiology, immunology, genetics, and biochemistry. Should one man’s beliefs negate the results of millions of scientific studies? The proposed mechanism of action of homeopathy is implausible, and therefore may be confidently discarded as anything more than a placebo. (And if you don’t believe me, Dr. Barker Bausell has a wonderful review of the homeopathy literature that confirms this).

Well, some of you plausibility skeptics say, Galileo’s beliefs (that the earth is round, not flat) were not accepted by scientists in his day (and yet he was right and they were wrong), so how can we say with certainty that any proposed mechanism of treatment is truly implausible? Isn’t anything possible?

I suppose a tornado could tear through a junk yard, creating a Boeing 747 by accident, though I’m pretty sure we’ll never see that on America’s Funniest Home Videos – though we may see it in a sci-fi movie. But I’m glad you asked about anything being possible – because this brings us to our final pillar of good science: reproducibility. If something is true, it should be true again and again, right? Galileo was proven right because in test after test, ships did not fall off the end of the “flat” earth. Its in the repetition of tests that we can confirm the truth. Next week we’ll take a final look at evidence through the lens of reproducibility.

Posted in: General, Homeopathy

Leave a Comment (24) ↓

24 thoughts on “Science-Based Medicine 101: Plausibility

  1. Scott says:

    Another important point about plausibility is that “implausible” does not mean “untrue.” It means “less likely to be true,” so it’s essentially a screening process rather than a robust conclusion.

    Most non-scientists’ objections to plausibility-based arguments are based on misunderstanding of this point, in my experience. It’s thought that saying “X is sufficiently implausible that it’s not worth the resources to do rigorous studies” is a smoke screen for “X is untrue”, when a good scientist will definitely mean precisely the former while taking no position on the latter.

  2. npgite says:

    Galileo’s heresy wasn’t that he said the earth was round instead of flat. Galileo was actually born after Magellan’s circumnavigation of the globe. He challenged the dogma that the earth was the center of the universe by claiming that it orbited the sun, along with the other known planets.

  3. qetzal says:

    But to believe that there is an actual physiological effect of shaken water for the treatment of all sorts of diseases, is to deny everything that is known about human physiology, immunology, genetics, and biochemistry.

    You can add pretty much all of chemistry and physics to that list as well.

    The Galileo part is garbled, though. Galileo’s controversial (but correct) belief was that the Earth orbited the Sun.

  4. seemstome says:

    “Galileo was proven right because in test after test, ships did not fall off the end of the “flat” earth. Its in the repetition of tests that we can confirm the truth.”

    Scientific research is not often as clear cut as tests for whether the earth is flat. In testing medical treatments, it seldom happens that a treatment either does or does not work, or that it has the same effect on every patient. In medical research, and in many other scientific areas, statistical analysis is required. And the results of statistical analysis are not always clear and simple to interpret. And different researchers do not usually get exactly the same results.

    So it is misleading to use the flat vs. round earth question as an example of how science should work.

  5. Ed Whitney says:

    Caution is always in order when using “placebo” as an explanatory principle. Homeopathy may exert its “placebo” effect by doing something that is actually effective. Homeopathic history-taking involves inquiring about many dimensions of the presenting condition: sensitivity to movement, relationship to hot and cold drinks, cravings for sweet or sour foods, exacerbation by anger, exacerbation by anxiety, by going from warm air to cold air or the reverse, etc. Homeopaths spend much more time with patients on average than do other practitioners, and much of this time is spent in pursuing the details of activity, mood, food, and other variables that affect the course of symptoms.

    This impossibly lengthy and detailed history taking may well have an effect on patients, who must stop and consider aspects of their complaints that had previously been unconscious. The “remedies” themselves may have nothing to do with the case. But the spirit in which the history is taken, the attention paid to apparently “trivial” factors that exacerbate and relieve symptoms, and the awareness that is brought to bear on the complaint, could have therapeutic effects that are worth considering.

    Now the plausibility of homeopathic practice shifts from violating the laws of chemistry and physics to the role of consciousness in healing. It may be objected that the responsive conditions are likely to be self-limiting, but most complaints that present to homeopaths have been present for considerable periods of time, and have not yet become limited. The questions in homeopathic history-taking often have to do with asking the patient, “How do you participate in the exacerbations of your condition?” Developing that awareness creates a biologically plausible explanation for the otherwise inexplicable. Avogadro’s number is irrelevant. The “ownership” of the course of the illness by the patient may be paramount.

  6. Joe says:

    @Ed Whitney on 13 Aug 2009 at 3:32 pm “Avogadro’s number is irrelevant. The “ownership” of the course of the illness by the patient may be paramount.”

    I don’t suppose you have any evidence for that.

    “May be” paramount, that is as powerful as “may not be” paramount. It’s like crossing a street- you have a 50% chance of surviving because either you get hit by a car or not.

  7. John Snyder says:

    @ Ed Whitney:

    As a pediatrician I have had many parents who have used, and reported improvements using, homeopathic medicines. I would estimate that less than 1% have ever seen a homeopath. The “medicines” are usually purchased over the counter (Oscillococcinum anyone?), and so the history taking was non-existant. Pure placebo effect, plain and simple.

  8. DVMKurmes says:

    I have seen the same phenomenon in my veterinary patients, with the same estimated exposure to homeopaths. If the homeopath is so important, why all the information and freely available remedies in nearly every whole foods, independent health food store, etc. all over the country. In the health food store near me, you can pick up dozens of homeopathic remedies complete with lists of symptoms for which they are supposed to be useful. The long and detailed history may certainly have a placebo effect of it’s own, but is not necessary for the effect that people think they see from the remedies.

  9. pmoran says:

    “Well, some of you plausibility skeptics say, Galileo’s beliefs ——were not accepted by scientists in his day (and yet he was right and they were wrong), so how can we say with certainty that any proposed mechanism of treatment is truly implausible? Isn’t anything possible?”

    Also the sun SEEMS to move around the earth. Galileo’s views were understandably counterintuitive for many in his day, imbued as they were also in religious beliefs that put man at the centre of the universe.

    Homeopathy lies at an even deeper level of implausibility, since no one anywhere has ever been able to enhance any property of a solute by diluting it.

  10. wertys says:

    Poor Galileo.

    He was tried for heresy because when he propounded his views about the heliocentric model, he did it as a dialogue between 2 proponents of the opposing views. Pope Urban was actually a supporter of Galileo but asked that the book contain his personal views in favour of geocentrism. Galileo obliged by calling the character propounding those arguments in his book Simplicio, and made him appear somewhat foolish and error-prone. Urban took this as personal caricature, and hence Galileo ended up fronting the Holy Office.

    People seem to like thinking that it was scientists and enemies who went after Galileo, but his downfall was brought about by a spat between former friends that came about b/c he couldn’t resist taking the piss out of his intellectual opponents. That will give you trouble when one of them is the most powerful political official in Europe !

  11. Ed Whitney says:

    As Thomas Aquinas says: Distinguo! (I may have accidentally deleted rather than posted a message to this effect.) There are three classical ways one may address a proposed premise: to concede, to deny, and to distinguish. When you distinguish, you both concede part of it, deny another part, and attempt to show that the part you have conceded does not negate your own premise.

    I distinguish between biological plausibility (the ostensible focus of this particular thread) and strength of evidence. I concede that I have no evidence for my premise, I deny the relevance of that lack to the present discussion, and assert that the lack of evidence (strength of effect, consistency of effect, and all those other Bradford Hill criteria) does not negate the plausibility that homeopathic history-taking may have a salutary effect.

    It violates no biological principles for a process of information-gathering to create a change in the relationship between a patient and his/her symptoms. If the patient does the inquiry himself rather than spending good money on a consultation, biological plausibility is not violated. The homeopathic books generally require the reader to plow through a large number of possible remedies, matching the reader’s symptoms to the descriptions in the book. The process of taking control of one’s condition, and becoming more acutely aware of exacerbating and alleviating factors, still has a plausible influence on the course of the complaints.

    I concede that there is a placebo effect, but deny that it is either pure, plain, or simple. As an explanatory principle, I find it wanting. I think I understand regression to the mean (a part of some of the alleged “effect” of placebo treatment). I have some notion of the power of suggestion on the severity of symptoms. But often patients go for homeopathy after conventional approaches have failed them. It is neither plain nor simple to understand why a placebo that comes out of a tube that says “Boiron” should succeed when a remedy with conventional pharmacological properties should have failed. This phenomenon is messy, deep, and complex.

    I distinguish between homeopathic practice (the entire process of inquiry, clinician-patient interaction, prescribing) and homeopathic “remedies” (those little white pills that come in those funny blue cylinders). Avogadro’s number is pertinent to the latter, but not to the former. I reassert the proposition that remedies have nothing to do with the case. I speculate (not assert) that some patients benefit from the sense of self-control that comes from taking charge of their own care.

  12. Harriet Hall says:

    (1) In Jay Shelton’s excellent book “Homeopathy: How it Really Works” he concludes that homeopathy works, but its effects have nothing to do with the remedies.
    (2) I keep wondering about those who think we shouldn’t dismiss anything on the grounds of implausibility – do they answer those e-mails from Nigeria?

  13. pmoran says:

    Ed: ” — assert that the lack of evidence (strength of effect, consistency of effect, and all those other Bradford Hill criteria) does not negate the plausibility that homeopathic history-taking may have a salutary effect. ”

    Who would argue, except, why select this one aspect of the homeopathic medical interaction? You quite rightly see any beneficial influences as not “pure, plain, or simple”.

    The likelihood and the strength of beneficial non-specific patient responses to medical care (including what are generally regarded as placebo “effects”) will depend upon an extremely wide range of factors to do with the therapist, the patient, the consultation and include the characteristics and credibility of anything prescribed.

    The placebo “treatment” will certainly often serve as a hook upon which the patient can hang a recovery for quite different reasons.

  14. Gabor Hrasko says:

    I think there is no real difference between what Ed Whitney and the others say. Just Ed emphasizes that homeopathy WORKS (meaning better than no treatment – though even this does not necessarily true in all cases), the others says that it is pure placebo, thus it DOES NOT WORK. In fact these two statements do not contradict; just we use two different meaning of “WORKING”.

    I think if we speak about placebo and about using placebo in medicine, we have to separate the topics:

    1/ Is the treatment works better than an adequate placebo? Is that placebo works better that no treatment?

    2/ Is the proposed mechanism plausible and justified?

    3/ Is it ethical and manageable to use placebos regularly in healthcare?

    In most of the homeopathy tests the placebo is selected in a way that it helps testing the “similia, dilution, succussion” theory (mechanism). The homeopaths care with the patients in the same way in both arms of the test. In fact in good tests they are also blinded.

    There are some setups that tests the “advanced care” theory and I guess they could be positive in several cases. But if one tells to a homeopath that his treatment is based upon the advanced care and not on Hahneman’s theory, he will definitely protest.

  15. Ed Whitney says:

    Personally I only like red placeobs. Those white ones don’t do a thing for me. And those large chalky ones make me feel worse.

  16. Ed Whitney says:

    With homeopathy, as with other forms of alternative medicine, there are often favorable outcomes attributable to unmeasured confounders. Patients who choose herbs, homeopathy, and other unorthodox interventions are more affluent, have more social capital, and have many lifestyle habits (tend to be nonsmokers, more physically active) associated with better health. There is a fair amount of literature to support this. Placebos “work” better if you fasten your seatbelt and take care of yourself in other ways. This describes the population that tends to use homeopathy. It is the practice pattern as a whole that confers biological plausibility on homeopathy and other forms of unconventional medicine.

  17. Gabor Hrasko says:

    “It is the practice pattern as a whole that confers biological plausibility on homeopathy”

    But this is a too wide definition of homeopathy. What you describe is the plausibility of the obvious theory that healthy lifestyle leads to more health. It has no relation to homeopathy. Using termiology in such a loose way you do is very much missleading.

    I am OK with your approach, if homeopaths agree that the “memory of water”, “laws of similars” and “potentisation” is a nonsense. But now in more and more science universities they start teachin these non plausible theories. And this could happen because they mix things the way you are doing. Your examples and the tests shows that the theatre around the basic concepts of homeopathy has some effect, but they also show that the basic concepts themselves are not correct (because the placebo pills are working just fine as well).

  18. John Snyder says:

    Ed:

    It is neither plain nor simple to understand why a placebo that comes out of a tube that says “Boiron” should succeed when a remedy with conventional pharmacological properties should have failed. This phenomenon is messy, deep, and complex….I speculate (not assert) that some patients benefit from the sense of self-control that comes from taking charge of their own care.

    By saying homeopathy is a placebo, plain and simple, I meant precisely that the placebo effect explains the reported “efficacy” of the treatment. I did not mean, as your specious comments infer, that the placebo effect is plain or simple. It is, obviously, quite a complex little devil. Nothing in your argument suggests anything other than a placebo effect, which includes (based on my understanding of the term) the individual’s sense of self-control to which you refer.

  19. anoopbal says:

    I guess the only limitation of the plausibility approach is it’s limited by what we currently know about the field. Sometimes it can be a problem in drug discovery and stuff.

    But biological plausibility is an excellent approach to weed out most of the CAM treatments.

  20. The Blind Watchmaker says:

    Please see the “Base Rate Fallacy” and Baysian Probability.

    These concepts are lost (or never introduced) to many studying science.

    Prior probability absolutely is a factor in formulating the plausibility of some concept.

    http://en.wikipedia.org/wiki/Base_rate_fallacy
    http://en.wikipedia.org/wiki/Bayesian_probability

  21. Be careful using “bayesian” in this context. One can understand statistics and not buy into subjective probability, or any of the other notions that “bayesian probability” might misleadingly suggest. The “Bayes formula” of course is something else: it is a theorem in most theories of probability in the uninterpreted calculus hence neutral to all interpretations.

    As for plausibility, one technique which I’ve been trying to pioneer is to use it ad hominem in the non-fallacious sense. For example, just say something like: “you don’t think that a mixture of car battery acid, mayonaise, silver powder and shredded saran wrap is a cure for cancer, right?” Then you can ask why not, etc. It isn’t even that your interlocutor has to think it isn’t; merely that it is *implausible* is enough. Plausibility (and its relative, consilience) help pare down what to test, since we cannot of course test everything in the universe in all combinations. It is vital to stress that this makes use of *background knowledge*, which is also largely tacit amongst scientists and technologists. Articulating the principles by which these inferences work is an open problem in the philosophy of science, one of which would have great benefits to investigate if more than a few truisms could be gathered.

  22. Charon says:

    wertys: no. Galileo had problems because he was trying to interpret the Bible in light of science, rather than the other way around. He was summoned before Cardinal Bellarmine several years before Urban VIII became pope.

    And what is with Galileo and the round Earth? Didn’t this come up before on SBM? WTF? For the record, famous people who proved the Earth was round include Aristotle (d. 322 BCE), Ptolemy (d. c. 168 CE), and Copernicus (d. 1543 CE). Galileo was born in 1564 CE.

Comments are closed.