PT is a bit like a cowboy movie complete with saloon etiquette, that can include the the virtual sounds of glass breaking. I became entangled lately in a thread on an orthopaedic forum, the name of which shall remain anonymous in order to spare it embarrassment, because I think it tries, it really tries to be science-based....
I have a few points to make about issues raised during the course of this thread, that directly relate to science and the PT version thereof, that speak to the role of being or at least striving to be a "clinical scientist".
First of all, PT "science" should align itself with and adhere to basic science 101 tenets. It shouldn't make up things for the sake of convenience, such as correlation in any way equaling causal relation. Yet in a PT textbook, apparently, there is being taught some mechanism for doing this very thing. Very pseudo-scientific IMHO.
Second, it should assume nothing, and operate according to the essential scientific principle that all hypotheses are there to be knocked over. (I.e., no one should take issue with others if their much beloved "hypothesis" and treatment construct curves around and smacks them in the back of their own head one day.)
Third, in the clinic we can literally BE scientific. E.g, an hypothesis exists that "pain comes from joints". No one seems to have tried to take that one down yet. Well, except for me and a handful of others perhaps. An easy way to take that one down, to disprove it, to thereby either improve or even disprove the "construct validity" of the "hypothesis", and by extension all treatments based on that hypothesis, would be to design a system that does not involve treating joints in any way. (One like mine, just as an example.)
Such a system must have construct validity based on something, of course, so let's pick updated pain theory. None of the variables in treatment should exist unaccounted for. As best I can, I have tried to account for all variables. I trust that pain science as developed by Patrick Wall and Ronald Melzack, and theorized as the neuromatrix model, have already weeded out confounding variables better than my own profession ever has, with all its vested interests, youth and lack of perspective. I am encouraged by the fact that prominent physiotherapists, such as David Butler, Lorimer Moseley, Louis Gifford and Michael Shacklock, have themselves taken on the task of examining neuromatrix theory, thinking through the ramifications, and have pioneered physiotherapy research pertaining to pain and our professional interface with it. They have designed methods to deal with pain that include constructs other than the ones based on "pain comes from joints", and their constructs may or may not exclude that particular hypothesis. (My own treatment construct deliberately DOES try to exclude it, because I believe it is a false hypothesis to begin with.)
One can't "prove" an hypothesis, one can only disprove one. I aim to disprove the "pain comes from joints" hypothesis, or in the attempt, at least refine it more.
So, with a treatment system geared at ONLY the surface of the body, and the nervous system, devoid of provocation testing of joints (because why include that if you are aiming to reduce pain instead of amplify it), or even of thinking about joints in any way at all, other than to test for range, I somehow managed to help someone (a young female figure skater) completely eliminate pain and dysfunction in her hip area in two visits; pain that was disabling (i.e., she was limping), that had been present for a month and had had other (unsuccessful and joint focused) practitioner involvement already. Furthermore, the girl in question went straight back to figure skate training full tilt again, with no problem, after visit number one.
To me, as a clinical scientist, this suggests that I managed, and am still managing, to disprove the hypothesis that "pain comes from joints", n=1, one at a time, one after another on through time.
Not only that, but when pain truly does come from a joint, as in the case of a woman I treated who had a stress fracture of the hip, it is easy to know that in fact the hip joint is involved when the parameters of the method are followed, and the patient responds outside the normal cluster of responses, i.e., does not respond, period, i.e., still can't weight bear without crutches.
After three attempts, time to send her back to the MD, even though the first x-ray prior to her ever seeing me was negative (i.e., they missed the fracture first time round). She in fact needed a pin. She came once more post op to mop up more pain she had from the whole traumatic process of the hip pinning. Once. Fine after that.
I think this process is what is involved in being a good responsible clinician with a grasp of basic science principles and a desire to get past all the clutter our profession saddles us with. Occam's chainsaw. Everyone can learn to use it.