To wear the mantle of Galileo, it is not enough to be persecuted: you must also be right.
–Robert Park
I used to spend a lot of time on the websites of Joe Mercola and Gary Null, the most influential medical cranks of the internets (to call them “quacks” would imply that they are real doctors, but bad ones—I will no longer dignify them with the title of “quack”). I’ve kept away from them for a while in the interest of preserving my sanity. Unfortunately, Orac reminded me this week of the level searingly stupid and dangerous idiocy presented by these woo-meisters.
In light of this, it seems reasonable to reexamine the Galileo gambit. When a “discoverer” of some new medical “miracle” is dismissed by the medical establishment, they often invoke the ghosts of Galileo and of Ignaz Semmelweis.
Galileo and Semmelweis are a pair of historical figures that share a common story—they both made significant scientific discoveries, documented the evidence for them, and were reviled by certain authorities, but eventually honored.
Ideas are cheap. I believe that my idea to use a flow sheet to track my diabetics’ care leads to better outcomes. I have precisely NO evidence to prove this, but it doesn’t harm me or my patients, and there is at least peripheral evidence elsewhere that this is a good idea. There is also a plausible hypothesis behind this—if I have one piece of paper that contains the critical data for a diabetic, I can see right away if their blood pressure or cholesterol are above optimal levels, I can see what their weight is doing, and I can see if they have engaged in proper preventative care, such as eye and foot exams. There is also a small body of data to support the practice. It would not surprise me if someone studies this in the future and finds my method lacking, especially vs. electronic health records. When necessary, I’ll happily modify my practice in a way that benefits my patients.
Let me summarize the characteristics of a “good” clinical science thinking, in this context (no, I’m not gonna go all Popper on y’all):
- Relevance: an idea should bear directly on a real clinical problem
- Testability: it should be possible to test the idea to see if it has merit (this includes Popperian falsifiability).
- Plausibility: the idea should have some basis in reality and should not have been birthed de novo from between someone’s buttocks. It should not require a “suspension of disbelief” or “open-mindedness”.
- Abandonability: the poser of the question should be willing to abandon the idea if it is proved false. Moving the goal posts, invoking a conspiracy, or any other deus ex machina is never necessary for a good idea.
- Modifiability: an idea can be rationally modified and retested if it may still contain a kernel of truth despite failing one or another tests. Any idea that is held so tightly that reality must be modified to fit the idea should be highly suspect.
There is an enormous literature on what constitutes science, etc. This is just a little guide to reading on quackery, crankery, and other idiocy.
When you encounter possible medical crankery, a couple of questions to ask yourself are “cui bono“: who benefits? Is the answer “patients”, “medical science”, or “one dude with a P.O. box”?
The other question is, “where’s the evidence?” (remember, no conspiracy theories or you violate Pal’s Law).
Or, as Dawkins so acerbically put it:
If you are in possession of this revolutionary secret of science, why not prove it and be hailed as the new Newton? Of course, we know the answer. You can’t do it. You are a fake.
Leave a Reply