Part one of this blog series gives historical context for Manual Therapy. Part two will dig into the research about what we’re doing and what we’re not doing with manual therapy and offer some ideas on how to clinically reason by using it.
The traditional MT approach (based in patho-anatomical and biomechanical theory) involves finding a lesion and specifically correcting that lesion. But are we able to do this? Can we rely on our hands and assessment skills to identify a movement anomaly at a specific segment?
This question brings us to what has historically been the guiding light of Manual Therapy: Motion Palpation. Otherwise known, in the spine anyway, as PPIVMS (Passive Physiological InterVertebral Movements) and PAIVMs (Passive Accessory InterVertebral Movements).
There has been a mountain of research on the reliability of motion palpation. Over 60 investigations (many annotated summaries here). The results have been mixed, but when reviewed and filtered for sound methodology the results overwhelmingly point to shockingly low reliability-- usually no or little better than what you would get from random chance. Check out this excellent review by Peter Huijbregts, which concludes that:
Intrarater (the same person at different time points) agreement varies from less than chance to generally moderate or substantial agreement.
Interrater agreement only rarely exceeds poor to fair agreement.
Rating scales measuring absence versus presence or magnitude of pain response yield higher agreement values than mobility rating scales.
Translation: we may be able to find pain, but we can’t reliably find hypo/hyper-mobile joints with motion palpation.
Next question, is motion palpation valid at testing segmental mobility?
Let’s start with what we think as a profession. 66% of polled PTs think PAIVMS are valid for assessing quantity of segmental motion, and 98% of manual therapists base treatment decisions at least in part on the results of segmental motion tests. Abbott et al.
These prevailing beliefs bump against the fact that motion palpation does not agree with motion measured by dynamic MRI (Landel et al). That is, there is no correlation between a manual therapist’s report of motion with a MRI machine’s findings of segmental motion. This begs a new question: what are we feeling when it clearly feels stiff? Could it be the inches of soft tissue between our hands and the joint? Perhaps the stiff areas are simply higher tone/viscosity muscle? Whatever it might be, we know that it's not the isolated motion of a single segment.
But let’s say that we could do it. I know many therapists who think that despite the evidence, they can feel this stuff. They acknowledge the studies and the methodologies and the fact that the therapists in these trials all had oodles of training, certifications, experience and expertise. But the traditionalists subvert science with an IME (In My Experience) claim, and boldly proclaim that research findings don’t apply to them. So let’s go with it, let’s say we have decided that a specific segment is hypomobile and needs some “adjusting”.
Well this brings us to another question: if we have a specific target identified, are we able to specifically mobilize and/or manipulate that segment with manual therapy?
Even with the best wind-ups, the best directions, and the best techniques, motion gets distributed over a broad area. In the majority of efforts the targeted segment doesn’t even get cavitated. Instead, we see the cavitations happening in other areas: the segments below, above, and even on the opposite side. (ref, ref)
Here comes the good news… it doesn’t matter!
Therapist selected MT interventions (based on assessment findings) have no advantage over randomly selected MT interventions for short term or for long term outcomes (ref, ref, ref). There is not a single trial out there that says otherwise.
The thing is that the effects of manual therapy are non-specific. They take place by way of a cascade of neuro-physiological events that affect the local area as well as the central nervous system--the Grand Pooba of pain, motor control, and psychology. To reiterate, the effects of MT are NOT from some sort of biomechanical or histological correction. Joel Biolosky says it better:
So how do these vague neurophysiological changes occur?
When we perform MT a mechanical force stimulates a whole whack of receptors and triggers a number of interesting things in a number of interesting ways. True, we don’t know exactly how it happens, but we do have a handle on some of the observable changes that take place after a dose of MT. Let's take a look at some of the work of Joel Bialosky:
Peripherally we see:
- A decrease in cytokines and substance p.
An increase in endorphins/cannabinoids (serum and blood stream).
Changes in electromyographic activity of local and regional muscles. We see high activity muscles decrease towards normal, and low activity muscles increase towards normal.
Spinally we see:
- Decreased temporal summation in the dorsal horn.
Centrally we see:
- Decreased activation in the pain centres (peri aqueductal grey, anterior cingulate cortex, and amygdala)
Changes in autonomics and in endogenous opioids
Changes in psychological features such as fear and expectation.
Interesting that there’s no mention of breaking scar tissue or adhesions. Could it be because you can’t do that without a scalpel? Or because there’s a frictionless skin-fascia interface that prevents forces from being directed in any direction besides perpendicular? Or maybe it’s because you need about 2000lbs of force to deform the IT band 1%. There’s also no mention of subluxations. I guess that’s because subluxations aren’t a thing. The literature in fact demonstrates that there is little to no mechanical effects for any manual treatment.
So where does this leave us? How should we go about applying manual therapy?
Don’t sink into the mires of worrying about identifying lesions and specifically fixing them, that’s a quick route to clinical angst and isn’t supported by anything.
In the words of the venerable Tim Flynn (a self proclaimed ‘recovering biomechanist’): “move it and move on”. This simple pragmatic approach takes the stress and preciousness out of MT; not to mention it’s much more valid.
Don’t tell your patient that you’re “fixing alignment”, “breaking up scar tissue”, or “putting something back into place”. Those are all lies and iatrogenically lead to dependence, lower self efficacy, and fragilistic thinking. AKA worse outcomes.
Optimize the placebo response (read this). Develop therapeutic relationship, foster expectations for a positive result, and bring to bear all the soft skills of rehab.
Use MT techniques as movement experiments. MWMs do this well and are excellent tools. MT can set up a novel experience for your patient and successfully decouple movement from pain. This framework fits well with an Exposure Therapy approach: set up movement experiments to violate an expectation of pain. MT is just one of many ways of doing this.
Use an asterisk sign (test-retest) approach to determine response and to demonstrate change to your patient. Use these changes as proof that their pain is malleable, “How could this be a serious/scary/structural problem if it can change with something so slight as a little pressure from my hands?”
- Use MT sparingly and only as a pathway to exercise and loading, which is the only tool we have to make real tissue changes.