What are the Medical Research Council grades for calculating the SAFE score?
The SAFE score is based on the MRC grades for shoulder abduction and finger extension of the paretic upper limb. Each movement is graded separately, and then the two scores added, to calculate the SAFE score.
You can use the grades below, and avoid using + and – signs to keep things simple.
0 = no palpable muscle activity
1 = palpable muscle activity, but no movement
2 = limited range of motion without gravity
3 = full range of motion against gravity, but not resistance
4 = full range of motion against gravity and resistance, but weaker than the other side
5 = normal power

Shoulder abduction should be tested with a long lever, resistance applied proximal to the elbow and taken through a full range of motion. Finger extension should be tested with the wrist stabilised, resistance applied distal to metacarpal phalangeal (MCP) joints and taken through a full range of motion.

If the fingers have unequal strength, use a majority rule. If three fingers have the same score, use this score. If two fingers have a lower score than the other two fingers, use the lower score. For patients with inattention or fatigue, effort needs to be made to ensure the patient is focused on the arm to give the most accurate indication of strength.

When is the best time to get the SAFE score?

We count the day of the stroke as day 0, the next day as day 1, and so on. It’s important to get the SAFE score each day until you are able to give them a prediction, or you get to Day 3, whichever occurs first. 
If a patient reaches 8/10 by day 3, then they are most likely to have an Excellent functional outcome for their upper limb. If a patient reaches 5/10 within 3 days, but is less than 8/10 at day 3, then they have an Excellent prediction if they are less than 80 years old, and a Good prediction if they are 80 years old or more.

Here are some examples:

Mr Paora, 87 y
Day 0 (day of symptom onset), SAFE = 6
Day 1 (day after symptom onset), SAFE = 8, so we can give him an Excellent prognosis right now, even though it’s only day 1. This is assuming he doesn’t deteriorate over the next couple of days, in which case we’d have to re-assess him. But assuming he’s stable, we expect him to remain at 8 or above, so he has he is most likely to have an Excellent functional outcome for his upper limb.

Mrs Peters, 74 y
Day 0 (day of symptom onset), unable to get a SAFE score, as she was admitted late that night and no staff were available
Day 1 (day after symptom onset), SAFE = 2
Day 2, SAFE = 4
Day 3, SAFE = 5. Now we can give her an Excellent prediction, as she has achieved a 5 on day 3, just in time, and is less than 80 years old.

Mr Lee, 63 y
Day 0 (day of symptom onset), unable to get a SAFE score, it was a Sunday and no staff were available
Day 1 (day after symptom onset), SAFE = 2
Day 2, SAFE = 3
Day 3, SAFE = 4, so we can’t give him an excellent prediction, and we remain unsure of his prediction if we don’t have TMS available. All we know for sure is that he’s not in the excellent category. 
If TMS was available we would obtain his NIHSS on Day 3, then wait to find our whether he was MEP+ or MEP- to determine if he was ‘Good’ (if MEP+), ‘Limited’ (if MEP- and NIHSS < 7), or ‘Poor’ (if MEP- and NIHSS ≥ 7). 

How should you score a patient when their overall strength is weaker than ‘normal’ (i.e. they would score 4/5 on the non-affected side)?
Score the affected arm relative to the non-affected arm. If the movement on the affected arm is weaker than the non-affected arm (but still has full range of movement against gravity and some resistance) score it as 4/5 for the SAFE score. If they can only move against gravity and not resistance, then they would score a 3/5. Discussing with a colleague or obtaining a second opinion is useful any time a SAFE score isn’t clear. Make your scoring rationale clear in any documentation.

How should you record a SAFE score when a patient has pre-existing pain, orthopaedic or musculoskeletal issues?
If the patient has a pre-existing biomechanical restraint or pain, you will need to perform a more thorough assessment to determine any influences from pain, passive joint restrictions or previous strength deficit that influences your current SAFE score.

Discussing with a colleague or obtaining a second opinion is useful any time a SAFE score isn’t clear. Make sure your scoring rationale is clear in the documentation.

When is the best time to get the NIHSS score?

If the patient’s SAFE score is less than 5 on day 3, get their NIHSS score straight away, and then organise the TMS test. This way, if they turn out to be MEP- you already have the NIHSS score from day 3, which is the best time to obtain it.

What is TMS?

TMS stands for Transcranial Magnetic Stimulation, which is a safe, painless, and non-invasive way of testing whether the motor pathways from the brain to the spinal cord are working. A brief magnetic stimulus is used to activate neurons in the motor cortex, which send a signal via descending motor pathways to muscles on the opposite side of the body. This generates a brief response in the muscles, called a Motor Evoked Potential (MEP), which can be recorded with surface electromyography. In PREP2 TMS is used to stimulate the motor cortex on the stroke-affected side of the brain, and see whether a message can get through to muscles of the paretic upper limb. If MEPs are observed (MEP+) this means that the motor cortex and its descending motor pathways are still functionally intact, even though the patient might have great difficulty activating the muscles themselves.

If a person had no responses to TMS 7 days after stroke, but responses returned at a later time, would they then have ‘good’ potential for hand and arm recovery?

If a person has no responses to TMS 7 days after stroke, this indicates a greater degree of stroke damage to the descending motor pathways. This means that even if they do recover responses later, they are unlikely to have the potential for a good upper limb functional outcome.

How important is the TMS? Can we skip this step and go straight to the NIHSS score?

The TMS step is important. We have found that people can have responses to TMS even though they have a SAFE score as low as zero, and an NIHSS score greater than 7. Being MEP+ shows that the remaining descending motor pathways are still functioning. This means that these patients still have potential for a good functional outcome. If the TMS step was skipped, the NIHSS score would put these patients in the limited or poor categories, which would underestimate their potential for recovery. If the pathways are functioning (MEP+), then it doesn’t seem to matter what their NIHSS score is.

What happens if the patient has another stroke during their admission?

A new stroke or progression of stroke symptoms which results in worsening upper limb weakness means that the PREP2 prediction tool should be started again. Count the day of the new stroke or progression as day 0. If the patient’s prediction category changes, you will need to convey the new prediction to the patient, their family, and the clinical team.   

Is it possible for the hand and arm to recover better than predicted?

Yes. The predictions made by PREP2 are accurate for 75% of patients. For the remaining 25%, about one-third have a better outcome than expected. Most of these are patients with a Good prediction who go on to have an Excellent outcome, and about 10% are people with a Poor prediction who do a little better than expected and have a Limited outcome. Note that patients with a Limited or Poor prediction are not able to have a Good or Excellent outcome, as the damage to their motor pathways prevents them from regaining fine motor control. It’s pleasing to see that some people do recover more hand and arm function than predicted by PREP2. This could be due to a range of reasons, which we are continuing to explore with our research. 

Can the hand and arm keep improving beyond 12 weeks after stroke?

Yes. The brain can continue to change in response to practice and experience, throughout life. Therefore it’s possible for improvements to be made beyond 12 weeks after stroke. However, most people make most of their recovery of movement within the first 12 weeks after stroke. This is because of the unique biological conditions in the brain during this time. Improvements can be made after this time, however they are usually smaller, slower to develop, and require more effort.

What type of upper limb therapy will ensure that people reach their predicted functional outcome?

PREP2 helps patients and therapists to focus on appropriate rehabilitation goals, but it can’t prescribe the type of upper limb therapy.

Why do some people not recover hand and arm function as well as predicted?

There may be a range of reasons why some people don’t reach their predicted upper limb functional outcome. For example, their function might be limited by problems with their cognition, attention, and communication. They may also have other health problems that make it difficult to engage in rehabilitation. We are continuing to explore these factors with our research.

Is referral for psychological support necessary when giving predictions?
It is important at the time of giving any PREP2 prognosis to be conscious of the effect this new information might have on a person’s expectations for their recovery.  Referral for psychological support should be considered for any patient who you believe is having emotional difficulties following a stroke. It is particularly important to consider this for those receiving the limited or poor prediction.