When families receive a stroke rehabilitation assessment report, the question carers ask most is: "The report says BBS 38 and TUG 18 seconds. What do these numbers say about the risk of falling?" This article takes a purely physiotherapy assessment view and goes through the balance and fall risk assessment tools commonly used in stroke rehabilitation: the Berg Balance Scale (BBS), Timed Up and Go (TUG), Functional Reach Test (FRT) and Single-Leg Stance (SLS). It explains what their scores mean and their evidence-based cutoffs, why one score is not enough, and when several tools are needed together.
This article aims to help families read the numbers on an assessment report. It is not for self-diagnosis and does not replace professional assessment. The causes of loss of balance after stroke, training strategies and the overall fall prevention framework belong to the parent topic of balance and fall prevention after stroke. This page concentrates on how to interpret the assessment tools and the logic of choosing them.
What balance assessment scores actually mean
The conclusion first. Although each of these balance scales has an accepted cutoff, no single tool can predict accurately, on its own, who will fall. Take the most commonly used tool, the BBS. A meta-analysis combining 21 studies (Park & Lee 2017) found that its overall performance in predicting falls is only "moderately accurate", with clear heterogeneity between populations and between cutoffs. The correct way to read it is therefore to treat the score as a starting point for clinical communication and for setting treatment goals, and for the therapist to make an overall judgement by combining several dimensions.
What does "moderately accurate" mean? An AUC of 0.84 means that the BBS does reasonably well in separating high-risk patients from low-risk patients in broad terms, but this does not mean it can decide exactly who will fall. Its sensitivity is 0.72 and its specificity is 0.73, which means a certain proportion of patients will still be put in the wrong group. When looking at a score, the timing of the test, the severity of the stroke and the other assessment tools must therefore be taken into account together.
The four commonly used balance assessment tools and their cutoffs
The dimensions measured by each tool differ. The list below first compares the evidence-based cutoffs and nature of the four commonly used tools, and each is then explained in more detail:
Assessment cutoffs are not absolute.The same BBS score can mean very different things in a patient who has just had a stroke and has severe hemiplegia and in a subacute patient about to leave hospital. The scores mentioned above are clinical reference ranges only. In actual interpretation, a physiotherapist must judge by combining the timing of the test, the severity of the stroke and the patient's overall condition.
Reference score ranges of each balance assessment tool
To help families form a basic idea of the numbers on an assessment report, the table below sets out the general reference ranges of the four tools above. It must be stated first that these values are ranges generally used for reference in clinical practice and are by no means hard diagnostic thresholds. Their real meaning differs from person to person and with the timing of the test. They are best treated as a starting point for discussion with the therapist, and a family should not match the patient to a category from the number alone.
| Berg Balance Scale (BBS, 0–56) | Clinically, a score of ≤29 is generally taken as a reference for a high risk of falling, and some stroke studies list a score of <45 as a range to watch. When the score gets closer to the full mark, watch for the ceiling effect, because a high score does not necessarily mean there is no risk of falling at all. |
| Timed Up and Go (TUG, seconds) | In stroke practice, about 14–15 seconds is generally taken as a reference line that needs attention. We suggest using the number of seconds to gauge the patient's mobility while also watching how steady the patient's gait is when walking, and the risk of falling should not be decided from the seconds alone. |
| Functional Reach Test (FRT, centimetres) | This test measures the limits of stability when the body reaches forward. Clinically, the result is generally compared with the last assessment to see whether there is progress, and the difference between the left and right sides is compared. It is better suited to tracking an individual's progress in recovery than to applying a fixed pass mark. |
| Single-Leg Stance (SLS, seconds) | This test measures static balance. It generally looks at whether the patient can stand steadily on one leg for a number of seconds, and compares the difference in performance between the affected and unaffected legs. Because it reflects only one side of balance, it needs to be looked at together with other dynamic and reactive tests. |
Why do we look at a reference range and not only at one fixed pass line? The same score means different things in patients of different severity and at different stages of recovery. In addition, the scales themselves have ceiling and floor effects, and applying one score line rigidly can easily lead to wrong judgements. The correct approach is to look at whether the score is improving, to use several test tools together, and to add the therapist's professional judgement, and not to stare at a single number.
Two limits of the BBS: ceiling effect and floor effect
The BBS is the most commonly used, but it has two structural blind spots that affect accuracy, and the score can only be read correctly when both are understood:
What is the ceiling effect (common in patients with better ability)
- When subacute stroke patients are about to leave hospital, many can already get the full score of 56/56
- The Inoue 2024 study showed that 43.1% of patients got the full score on the BBS, but on the more difficult Mini-BESTest only 5.2% got the full score
- A full score does not necessarily mean no risk of falling at all; if the patient is a little unsteady when moving about or when suddenly losing balance, the BBS may not be able to detect it
What is the floor effect (common in patients who have just had a severe stroke)
- Most patients who have just had a stroke with severe hemiplegia get very low scores, all crowded at the bottom end of the scale
- The harder tasks in the scale cannot separate the small differences between these low-scoring patients
- So the same score in the acute phase and in the subacute phase means something different clinically
In short, the BBS discriminates best in the middle of the score range, and tends to bunch scores at the high and low ends where they are hard to separate. A patient who scores 56 in the subacute phase does not necessarily have flawless dynamic balance. In the same way, a patient who scores low in the acute phase may not be separated by the scale from another low-scoring patient in terms of ability. This is why physiotherapists do not rely on one BBS number alone.
When several assessment tools are needed together
Since each tool has its own range of measurement and its own blind spots, a physiotherapist will combine different tools according to the patient's real ability, to fill in the dimensions the BBS lacks, especially reactive and anticipatory postural control and sensory integration:
- The BBS easily shows a ceiling effect and has difficulty catching subtle dynamic instability
- Add the Mini-BESTest: it covers reactive and anticipatory postural control and sensory integration, with less ceiling effect (Inoue 2024: full score only 5.2%)
- Or add gait speed and the TUG, to improve discrimination of falls during dynamic movement
- BBS scores bunch at the low end, and the harder tasks in the upper half are hard to separate
- Focus on more basic tests such as sitting and trunk control, together with clinical observation
As for why the Mini-BESTest can make up for the BBS, the prospective cohort study by Inoue 2024 found that in subacute stroke patients the AUC of the two for discriminating falls and walking ability is in fact quite close (BBS about 0.72–0.73, Mini-BESTest about 0.74), but the Mini-BESTest has better responsiveness and less ceiling effect. For high-functioning patients it picks up recovery progress and subtle instability more sensitively. In other words, the value of using several tools together is that they cover each other's blind spots.
Do not confuse standing and dynamic balance tests with sitting trunk assessment. The BBS, TUG, FRT and SLS introduced on this page are all used to assess the patient's overall balance and fall risk when standing and moving. Another tool, the Trunk Impairment Scale (TIS), assesses the balance and coordination of the lower back and trunk itself while the patient is sitting. This is a more basic kind of body control, and is usually used when the patient cannot yet stand steadily or walk. How to read the TIS score, and what the cutoff of 12 means for predicting the course of recovery, is a separate topic. The two kinds of assessment have different uses and do not overlap.
Another important part of balance: sensory integration
Balance does not depend on muscle strength alone. Impaired proprioception and sensory integration after stroke directly affect steadiness in standing and increase the risk of falls, although the BBS or TUG alone may not fully reflect this dimension. If during the assessment the patient is found to do clearly worse with eyes closed, when stepping on uneven ground or when facing visual distraction, the therapist takes the sensory dimension into account and arranges targeted training.
A reminder about the division of professional roles: This page mainly covers the balance and mobility tests commonly used in physiotherapy. How cognitive problems and visuospatial neglect after stroke affect the risk of falling, and the safety assessment of the patient in the home living environment, belong to occupational therapy; see occupational therapy after stroke. In real rehabilitation, physiotherapy and occupational therapy are best used together, and this page does not go into that part in detail.
How often should balance be reassessed
Families often ask: "After one assessment, when does it need to be done again?" In fact a single score only reflects which stage the patient is at now, and the trend of progress can be seen only by repeat assessment, and the trend is often more useful than one absolute score. The following is general practical advice, and the actual rhythm of reassessment still has to be adjusted flexibly by the therapist according to the patient's phase of recovery:
At a clinic or during regular treatment
- It is usually advised to test again with the same assessment tool after a period of training, so that scores can be compared under the same standard to see whether there is progress
- Patients recover faster from the acute phase to the subacute phase, so reassessment is usually more frequent, which lets the therapist adjust the training goals in time
- At each reassessment, use the same scale and the same test method as far as possible, so that the change in score before and after is accurate
After discharge or during rehabilitation at home
- In general, if the patient's physical condition changes clearly (for example a recent fall, loss of strength, a change of medicine or a change in the home environment), contact the therapist to arrange a reassessment
- Family members can also pay attention to how the patient manages at home, for example whether getting up from a chair, turning or stepping over a threshold is harder than before, to judge whether another assessment is needed
- Daily observation by the family can serve as a prompt, but a formal balance assessment score still has to be tested and interpreted by a physiotherapist
Limits of stroke fall risk assessment and how to use it correctly
To use these assessment tools properly, the following limits must be recognised at the same time:
Limits of the assessment tools themselves
- No test tool can predict with 100% accuracy who will fall; for example the overall performance of the BBS is only moderately accurate (AUC 0.84, Park & Lee 2017)
- The result of the TUG itself is very stable, but when it is used alone to predict falls in people with stroke the research results are not very consistent, so it is better suited to measuring basic mobility (Hafsteinsdóttir 2014)
- Because of the ceiling and floor effects, the same score means different things at different test times and in patients of different severity
How to view and use assessment scores correctly
- Assessment scores are mainly used to help healthcare staff communicate and set training goals, and are not a tool for the patient or family to diagnose themselves
- When looking at a score, the timing of the test, the severity of the stroke and the patient's overall condition must be taken into account together
- How cognitive problems or visuospatial neglect affect the risk of falling belongs to occupational therapy (OT) assessment, and this page does not discuss it in detail
Exercise prescriptions for fall prevention aimed at the general public and at community level (not specific to stroke) are a separate topic. Training programmes by stroke severity are part of home stroke rehabilitation. This page concentrates on assessment methodology that applies across severities and does not repeat specific training prescriptions.
- Treat the assessment score as a starting reference, and understand it together with the timing of the test, the severity of the stroke and the other test results
- Note that even a full BBS score (56) does not necessarily mean there is no risk of falling, because it may be affected by the ceiling effect
- For patients with better mobility, it is best to do the Mini-BESTest and a walking speed test as well, to add an assessment of reactive ability and dynamic balance
- Do not decide whether the patient's risk of falling is high or low from one BBS score or the seconds of one TUG alone
- Do not look up assessment scales online to diagnose the patient yourself, as this cannot replace a physiotherapist's professional assessment