Evidence-based physiotherapy and rehabilitation resources
Editorial & review policy
Evidence-based physiotherapy and rehabilitation resources
Editorial & review policyNeurological assessment
Pick up a patient's arm and move it. Something pushes back. Learning to tell what is pushing back — and how to put a number on it that another clinician can repeat — is one of the genuinely difficult skills in neurological physiotherapy. This is how it is done.
Imagine two patients. Both have had a stroke. You lift each one's forearm and straighten the elbow, and in both cases you meet resistance at roughly the same point.
It would be natural to write the same thing in both charts. It would also be wrong, because the resistance in those two arms may be coming from completely different places — and the treatment that helps one will do nothing at all for the other.
So before any scale or any number, get this idea straight: the resistance you feel is a mixture. Some of it comes from the muscle and its surrounding tissue simply being what it is. Some of it comes from the nervous system actively pulling against you.
Figure 1 · Where the resistance comes from
The three on the left of that diagram are properties of the tissue. Muscle and its connective tissue have springiness and thickness, and if a muscle has been sitting short for months, it becomes short. None of that involves a nerve firing. If you anaesthetised the patient completely, all three would still be there. [3]
The two on the right are the nervous system. One of them — only one — is spasticity. [3]
The single most useful thing on this page
Resistance from the tissue is called stiffness. Resistance from the reflex is called spasticity. The whole thing together, which is what your hand actually feels, is called hyper-resistance. Thirty-seven specialists across twelve countries agreed on those words precisely because everyone had been using them to mean different things. [3]
Get in the habit now of saying "increased resistance" until you know which kind it is. It sounds cautious. It is simply accurate.
Here is the part that makes everything else make sense.
Your muscles contain their own stretch detectors. When a muscle is pulled, those detectors fire, a signal runs into the spinal cord, and the cord sends a signal straight back telling the muscle to contract. That loop is fast and automatic. It is happening constantly, and it is why your knee jerks when it is tapped.
Normally the brain sends a steady damping signal down the cord that keeps this loop quiet. Think of it as a brake that is always lightly applied.
Figure 2 · The loop, and the brake on it
Now damage the pathways carrying that brake — a stroke, a cord injury, cerebral palsy. The loop is intact. The brake is not. So the loop starts answering back much more strongly than it should. [1]
And crucially, it answers in proportion to how fast you pull. Move the limb slowly and the detectors barely notice. Move it quickly and they fire hard, and the muscle grabs.
The definition, and why it is worded that way
Spasticity was defined in 1980 as a speed-dependent increase in the stretch reflex, from a reflex that has become too easily triggered. [2] Every clinical test that follows is really just an attempt to catch that speed-dependence with your hands.
This also explains something students find confusing. A stiff, shortened muscle resists you the same amount whether you move fast or slow — because tissue does not care about speed. A spastic muscle resists you far more when you move fast. The difference between fast and slow is the signal you are hunting for.
This is the scale you will see used on almost every ward round in the country. It is quick, it needs no equipment, and everybody understands the number. [5]
Figure 3 · The six grades in plain words
Two things about that ladder are worth pausing on.
First, notice 1+. It was inserted between 1 and 2 to capture something clinicians kept meeting, and it is why the scale is a set of ordered steps rather than real numbers. A grade of 4 is not twice a grade of 2. You cannot average these across a limb, and an averaged tone score is a number that does not mean anything. [5]
Second, the scale was introduced on the back of a study of thirty patients, testing only the elbow. The two researchers agreed with each other most of the time, and they were careful to say their results applied to that one muscle group. [4] We now use it on every muscle in the body. That is not their fault, but it is worth knowing.
What this scale cannot do
Look back at the first diagram. The Modified Ashworth grade is one number for the whole bar. It cannot separate tissue from reflex. [5] If your patient's muscle has quietly shortened over three months, their grade goes up without a single thing changing in their nervous system.
The Tardieu Scale exists because of exactly that problem. Instead of moving the limb once and grading the feel, you move it twice, at two different speeds, and compare. [5]
Figure 4 · R1, R2 and the gap
Two practical tips nobody writes down
Standardise your fast. "As fast as you can" is different on a Monday morning and a Friday evening. One group improved their results simply by setting a metronome and moving in time with it. [6] A phone metronome costs nothing.
Sometimes there is no catch at all. If nothing grabs, there is no R1 to measure. Write that down as a finding. Do not invent an angle to fill the box. [5]
Now the two tests start doing real work. You have two pieces of information: how much range the joint has, and how big the gap is between slow and fast.
Figure 5 · Putting the two findings together
Go back to the two patients from the beginning. Both graded the same on the Ashworth scale. But if the first has a wide gap and near-normal range, and the second has a narrow gap and a joint that stops early, you are looking at two different problems wearing the same number. [5] One needs the reflex addressed. The other has a tissue that has grown short, and no amount of treating the reflex will lengthen it.
There is a further method that pushes this idea further still, measuring five separate angles rather than two, so that shortening, spasticity, weakness and tiring-out can each be given their own value. [7] You are unlikely to meet it early on. It is worth knowing it exists, because it shows where careful assessment can go.
Honestly? Less than you would like, and this is not a reason to despair.
When researchers pooled every study they could find on the Modified Ashworth Scale, two patterns came out clearly. You agree with yourself much better than you agree with a colleague. And everybody does better in the arm than in the leg — which is unsurprising, since the scale was built on the elbow. [10] In the leg, two clinicians grading the same patient disagree often enough that a one-grade difference between you means very little on its own.
| If you want to… | Then do this |
|---|---|
| Track whether a patient is changing | The same person takes both measurements, in the same position, at the same speed [5][10] |
| Hand over to a colleague | Give them your position, your speed and your raw angles — not just the grade [5] |
| Report a Tardieu result | Report R1 and R2 separately, not only the gap. The gap is the least reliable of the three, because it carries the error of both. [6] |
| Make a decision about one patient | Be careful. These scales behave better across groups than they do for any single person. [5] |
You should also know that specialists do not fully agree about these scales. An international panel that reviewed every available measure recommended the Tardieu and set the Ashworth family aside, on the grounds that it measures tone rather than spasticity. [9] A separate group has argued that the Tardieu is not well enough tested either. [2] A systematic review sat between them, finding the evidence neither strong enough to endorse nor poor enough to reject. [8] Newer technologies are being developed, and are not yet ready to replace your hands. [11]
None of that means the tests are useless. It means you should describe what you did and what you felt, and be modest about what one number proves.
A good note lets someone else repeat exactly what you did. Each line below exists because leaving it out is a known reason two clinicians end up disagreeing.
| Write down | Because |
|---|---|
| Which muscle, which joint, patient position | Every reliable study fixes these first [5] |
| Which scale you used | Ashworth, Modified Ashworth and Tardieu are not the same test [5] |
| How fast you moved, and how you controlled it | Speed changes the answer [2][6] |
| For Tardieu: R1, R2 and the gap, listed separately | The gap is the least reliable of the three [6] |
| Whether you felt a catch at all | No catch means there is no angle to record [5] |
| Your own name | Because the next reading should ideally be yours too [10] |
| Any shortening you noticed | It pushes the grade up on its own [5] |
Two stations built only from the method above. Every step is one you can justify if an examiner asks why you did it.
Practice station 1 — Modified Ashworth, elbow flexors
Task. Assess tone in the elbow flexors of a patient after stroke, and report what you find.
Required sequence
Presentation. Say the grade, the muscle, the position you used and the speed you used. Describe it as resistance to movement rather than as spasticity, because this test cannot tell the two apart. [5]
Common errors
Practice station 2 — Tardieu at the ankle
Task. Assess the calf muscles with the Tardieu Scale and explain what your two numbers mean.
Required sequence
Presentation. Report R1, R2 and the gap separately, and say how you controlled your speed. A wide gap points to the nervous system; a small gap with a restricted R2 points to the tissue. [5]
Common errors
The scenarios below are made up for teaching. The patients are not real, and everything in them illustrates something set out earlier on this page.
Two patients, one grade
Both patients score 2 on the Modified Ashworth Scale at the ankle. You go on to do a Tardieu on each.
| Finding | Result |
|---|---|
| Patient A, slow movement (R2) | Close to a normal range |
| Patient A, fast movement (R1) | Stops much earlier — a wide gap |
| Patient A, what you felt | A clear catch |
| Patient B, slow movement (R2) | Stops well short of normal |
| Patient B, fast movement (R1) | Almost the same as R2 — a narrow gap |
| Patient B, what you felt | No catch at all |
What it means. Patient A's ankle stops far earlier when you move fast, so the restriction depends on speed — that is the reflex. [5] Patient B's ankle stops at much the same place either way, and the range itself is small, which points to a muscle that has grown short. [3] The Ashworth grade was identical and told you none of this.
What to do with it. Write the two up differently, in words as well as numbers. For patient B, note that there may be no angle of catch to record at all, and that is a finding rather than a gap in your assessment. [5]
Your colleague got a different answer
You grade a patient's knee as 2 in the morning. Your colleague grades the same knee as 3 after lunch. The team asks which of you is correct.
| Finding | Result |
|---|---|
| Which limb | Lower — where this scale performs worst [10] |
| Did either of you standardise the speed? | No |
| Was the position recorded? | No |
| Same rater both times? | No |
What it means. Neither of you is wrong. Two different people grading the same leg disagree often, and the scale is known to do worse in the leg than the arm. [10] The question the team has asked cannot be answered as it stands.
What to do with it. Agree a position and a speed, write them in the notes, and have the same person take the repeat measurement. [5][6] Then a change of one grade starts to mean something.
By far the commonest error, and it makes spasticity vanish. The reflex only answers to speed. [2] If you move gently and politely, you will feel tissue and nothing else.
The Ashworth grade cannot tell you which kind of resistance you met. [5] Write what you felt, not what you assume caused it.
A patient can improve neurologically and score worse, because their muscle has been getting shorter while you were not measuring it. [3] Without the slow measurement you will never see this happening.
Six ordered steps, one of which is called "1+". [5] Averaging them produces a number with no meaning.
In the lower limb especially, disagreement is the expected behaviour of the tool, not a failure by either of you. [10] Standardise the method instead of debating the result.
Weakness and loss of control are usually the bigger cause of lost function, and no tone scale detects them. [1] The quiet, floppy arm can be the more disabled one.
Ten questions on what you have just read. Tap one to see the answer and why.
Answer: (B) The reflex responds to speed; tissue does not. Resistance that is unchanged by speed points to stiffness rather than spasticity. [3]
Answer: (C) The loop itself is intact. What is lost is the descending control that normally keeps it quiet, so the loop answers a stretch far more strongly than it should. [1]
Answer: (B) From the shortest position to the fullest stretch in roughly one second. Moving more slowly is the commonest reason a genuine catch is missed. [5]
Answer: (B) R2 is measured during the slow movement and is the range available. R1 is the angle of catch during the fast movement. [5][6]
Answer: (B) The gap is the part of the restriction that depends on how fast you move, which is the part the reflex is causing. [5]
Answer: (C) Both angles are measured reasonably consistently, but the difference between them carries the error of both, and it did not reach an acceptable level even with the speed standardised. [6]
Answer: (C) Shortening is a tissue contribution to the resistance. It pushes the grade up without any change in the nervous system, and a single-number scale cannot separate the two. [3][5]
Answer: (B) The six grades run 0, 1, 1+, 2, 3, 4. They tell you the order of severity but not the size of the gaps between the steps, so arithmetic on them produces a meaningless number. [5]
Answer: (C) With no catch there is no angle of catch to measure. Record the absence — it is a finding, not a failed test. [5]
Answer: (C) Disagreement between raters is expected with this scale, and it is worse in the leg than the arm. Standardising the method and using the same rater is what makes a change meaningful. [10]
Everything on this page, in one screen
Reviewed by the Physiotherapist India Team.
How to use this
It sets out what is actually happening in the limb, then the two tests you will be asked to perform, with diagrams for the parts that are hard to picture. Faculty may use this page in teaching with attribution. It carries its review date and its next review date, so you can see at a glance whether it is current before you put it in front of a cohort.