Apr 20 2009
Hunch-Based Medicine
Last night I listened to the latest edition of the The Skeptics’ Guide to the Universe podcast. In one segment the crew discussed the recent study/meta-analysis of a homeopathic remedy I blogged about a couple days ago [Imperfect Research Into Homeopathic Remedies]. Much of their criticisms and concerns echoed mine. At the end of the segment the show host, Steven Novella, asserted, in so many words, that emphasizing evidence-based medicine is not sufficient. Rather, the issue should be whether or not medicine is science-based.
What’s the difference? As I see it, science-based medicine bases its hunches and hypotheses and theories upon a pre-existing body of scientific knowledge. The ideas and explanations that provide the superstructure for the data — if you will — rests upon the solid foundation of what we already know about human physiology and chemistry, etc. Not strictly and completely, but at least substantially.
Over the past few years there has been a trend for alternative treatments to start gathering data. Evidence. Which is good. Data is the most important element to doing science. But there is more to science than data.
In the past and still today there was/is precious little data to support the majority of claims about alternative treatments. Most of the data was of very poor quality — anecdotes and testimonials and trials lacking adequate controls. But that is changing. Which is good. If a treatment works, we should verify it does. Or in many if not most cases, if it doesn’t.
But there is more to medicine than that, if it is to be a science. In science we have data, but not just data. There are hypotheses and theories. Hypotheses will propose a relationship between variables. Theories will offer testable explanations of some class of events. A strong theory will reveal the actual real-world mechanisms the produce the phenomena in question. Testable mechanisms.
While the one study I criticized did provide some data, and hence partly qualified as science, the entire field of homeopathy fails in another regard: the theoretical superstructure of their “knowledge” appears to be bogus. For one, no principles or tests have shown that diluting a medicine actually makes it stronger. In fact, the opposite is true. And as for the like-curing-like idea central to homeopathy (the homo, “same” part), that, too, appears to be bogus.
So yes, encourage and applaud the gathering of data. But there is more to a robust understanding than that. But, in terms of combating pseudoscience and sham treatments, we have to start somewhere. So yes, start with tests. Good tests.
Proponents of “alternative medicine” may accuse we skeptics of “moving the goalposts” by asking for more than data that some specific treatment works better than placebo. But intellectual progress depends upon it. In a sense, by scrutinizing both the effectiveness AND supposed mechanisms of action, we are saying, “let’s take a closer look at this.”
Can taking a closer look be a bad thing?
