1013

Reliability of the Tone Assessment Scale and the Modified

Ashworth Scale as Clinical Tools for Assessing
Poststroke Spasticity
Janine M. Gregson, MRCP, Michael Leathley, PhD, A. Peter Moore, MD, Anil K. Sharma, FRCP,
Tudor L. Smith, MCSP, Caroline L. Watkins, BA(Hons)
ABSTRACT. Gregson JM, Leathley M, Moore AP, Sharma
AK, Smith TL, Watldns CL. Reliability of the Tone Assessment
Scale and modified Ashworth scale as clinical tools for assessing poststroke spasticity. Arch Phys Med Rehabil 1999;80:
1013-6.

Objectives: To establish reliability of the Tone Assessment

Scale and modified Ashworth scale in acute stroke patients.
Setting: A North Liverpool university hospital.
Patients: Eighteen men and 14 women admitted with acute
stroke and still in hospital at the study start date (median age, 74
yrs; median Barthel score, 8).
Main Outcome Measures: The modified Ashworth scale
and the Tone Assessment Scale.
Study Design: The 32 patients were examined with both
scales on the same occasion by two raters (interrater comparison) and on two occasions by one rater (intrarater comparison).
Results: The reliability of the modified Ashworth scale was
very good (Kw = .84 for interrater and .83 for intrarater
comparisons). The reliability of the Tone Assessment Scale was
not as strong as the modified Ashworth scale, with marked
variability in the assessment of posture (K = .22 to .50 for
interrater and .29 to .55 for intrarater comparisons) and
associated reaction (K/Kw = --.05 to .79 for interrater and. 19 to
.83 for intrarater comparisons). However, those aspects of the
Tone Assessment Scale that addressed response to passive
movement and that are scored similarly to the modified
Ashworth scale showed good to very good interrater reliability
(Kw = .79 to .92) and good to very good intrarater reliability
(Kw = .72 to .86), except for the question related to movement
at the ankle where agreement was only moderate (Kw = .59).
Conclusions: The modified Ashworth scale is reliable. The
section of the Tone Assessment Scale relating to response to passive
movement is reliable at various joints, except the ankle. It may assist
in studies on the prevalence of spasticity after stroke and the
relationship between tone and function. Further development of a
measure of spasticity at the ankle is required. The Tone Assessment
Scale is not reliable for measuring posture and associated reactions.
1999 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and
Rehabilitation
From the Stroke Team for Audit and Research, University Hospital, Aintree, and the
Departments of Nursing and Neurology, University of Liverpool, Liverpool, United
Kingdom.
Submitted for publication September 1, 1998. Accepted in revised form February
16, 1999.
Supported by the Aintree Stroke Unit Trust Fund.
No commercial party having a direct financial interest in the results of the research
supporting this article has or will confer a benefit upon the authors or upon any
organization with which the authors are associated.
Reprint requests to Dr. Janine M. Gregson, Stroke Team for Audit and Research,
North Liverpool University Hospital at Aintree, Longmoor Lane, Liverpool L9 7AL,
United Kingdom.
1999 by the American Congress of Rehabilitation Medicine and the American
Academy of Physical Medicine and Rehabilitation
0003-9993/99/8009-518853.00/0

PASTICITY IS WELL recognized after stroke ~ and is

characterized by increased muscle tone with exaggerated
tendon jerks.2 A survey has found that although 94.4% of health
care professionals consider spasticity to be of clinical importance, it is rarely measured and quantified) Together with loss
of motor power and disordered sensation and perception,
spasticity is commonly thought to contribute to poststroke
functional impairment. Although anecdotal clinical experience
suggests that this is the case, no empirical evidence is available.
To establish the relationship between abnormal tone and
function, it is important to examine both in the same patient at
the same time. Therefore, although a valid and reliable measure
of function exists (the Barthel activities of daily living index4),
a valid and reliable measure of tone is still needed. Although
there is no direct method of measuring spasticity, the modified
Ashworth scale s (MAS) is the most frequently cited of the
available clinical rating scales to measure tonal abnormality. 6 It
is already being used to evaluate the effects of drug treatments
for spasticity. 7
Despite the accepted use of the MAS, there are no specific
written guidelines for standardizing its use. Furthermore, reliability has only been demonstrated for its use in measuring
spasticity at the elbow s and at the wrist. 8) These studies,
however, are flawed methodologically: Bohannon and Smith s
used ill-defined inclusion criteria and excluded patients with
cognitive impairment, and Bodin and Morris 8 excluded patients
with both cognitive impairment and with "known" spasticity.
This has implications in that if the MAS is not reliable, prior
exclusion of patients thought to have no spasticity would skew
the data analysis if in fact some of these patients did have
spasticity. The MAS takes no account of the relation of
abnormal tone with posture 9 and associated reaction, 1,11 both
of which may be important the measurement of tone and its
impact on function. A Tone Assessment Scale (TAS) has been
developed in an attempt to address these issues (unpublished
observations). It has written guidelines and contains items
relating to both posture and associated reaction, but a pilot
study did not find reliability for all items, possibly because of
small sample size and/or limited training of raters (unpublished
observations).
To study the relation between abnormal tone and impaired
function, it is first necessary to demonstrate a reliable measure
of tone. This study (1) reevaluated the TAS using a larger
sample size and increased rater training, (2) reevaluated the
MAS at the elbow using an adequate sample size and standardized written guidelines, and (3) compared and contrasted the
potential usefulness of the TAS and MAS as clinical and
research assessment tools.

SUBJECTS AND SAMPLING

All consenting patients with a clinical diagnoses of acute
stroke and were still in hospital at the study start date were
Arch Phys Med Rehabil Vol 80, September 1999

1014

MEASURES OF SPASTICITY, Gregson

Table 3: Order of Group Assessment

Table 1: Tone Assessment Scale

0
1.
2.
3.
4.

5.

6.
7.

8.

9.

10.

11.
12.

Isthe hand resting on the leg?

Are the shoulders level?
Is the foot flat on the floor?
Can you straighten the fingers, with
the forearm in midposition and the
wrist extended (sitting)?
Can you flex the hand to the mouth
and then fully extend the elbow
within 2 seconds (sitting)?
Is the lower limb flexible and the
knee easily extended in sitting?
Can you dorsiflex the foot from 20
to 10 of plantarflexion (mid-rotation, leg extended, patient supine)?
Can you passively flex the hip/knee
to 90 and return to full extension
within 2 seconds
Can you flex the knee with the hip
extended, to move the foot over the
edge of the bed, without resistance?
Does the hand remain stationary on
the leg as the subject elevates the
opposite arm above the head?
Can the hand remain at trochanter
level or lower on standing up?
Can the foot remain on the floor on
standing up?

considered for inclusion. Diagnosis was confirmed by computed tomography (CT) in all patients. There were no age
restrictions. Exclusion criteria included nonconsent, inability to
comply with the conditions of the standardized written guidelines, and prescription of tone-modifying drugs. No presumptions were made about patients with "known" spasticity,
because specific inclusion or exclusion of such patients would
have biased the data.

Methodology
The following demographic and stroke-related variables
were recorded: age, gender, time poststroke, type of stroke
(ischemic versus primary hemorrhagic), Oxford Community
Stroke Project (OCSP) 12 classification of stroke, and previous
history of stroke or transient ischemic attack. Barthel score 4
was measured at postadmission day 7.
Table 2: Modified Ashworth Score 5
Grade

Description

0
1

No increase in muscle tone.

Slight increase in muscle tone, manifested by a catch and
release or by minimal resistance at the end of the range of
motion when the affected part(s) is(are) moved in flexion
or extension.
Slight increase in muscle tone, manifested by a catch followed by minimal resistance through the remainder of the
range of motion but the affected part(s) is(are) easily
moved.
More marked increase in muscle tone through most of the
range of movement, but affected part(s) easily moved.
Considerable increase in muscle tone, passive movement
difficult.
Affected part(s) is(are) rigid in flexion or extension.

3
4
5

Arch Phys Med Rehabil Vol 80, September 1999

Rater 1
Rater 2

Day 1

Day 2

Day 3

Day 4

Group A
Group B

Group B
Group A

Group A

Group B

The standardized written guidelines that already exist for the

TAS (table 1) were reviewed and revised as follows.
1. Posturing at Rest (Q1-Q3): Score 0 if YES and 1 if NO.
2. Response to Passive Movement (Q4-Q9): Score 0 if there
is no increase in muscle tone; score 1 if there is slight increase
in muscle tone, manifested by a catch and release, or by
minimal resistance at the end of the range of motion (ROM);
score 2 if there is slight increase in muscle tone manifested by a
catch, followed by minimal resistance throughout the remainder
(less than halt) of the ROM; score 3 if there is more marked
increase in muscle tone through most of the ROM, but the
affected part is still easily moved; score 4 if there is considerable increase in muscle tone, passive movement is difficult
(cannot move through range in time indicated); score 5 if the
affected part is rigid in flexion or extension.
3. Associated reaction, Q10 and Q l l : 1 = less than 30 of
elbow flexion; 2 = 30 to 50 of elbow flexion; 3 = less than
50 of elbow flexion. Q12: Score 0 if YES and 1 if NO.
Standardized written guidelines for the MAS (table 2) were
developed as follows.
Test area: The testing area is quiet and screened from other
patients and therapists.
Starting position: Each patient is positioned supine on a
treatment couch (or as close to supine as the patient is able to
tolerate comfortably).
Testing Procedure: The forearm is grasped distally (just
proximal to the wrist); the forearm is in neutral supination. The
arm is moved passively by the examiner and is stabilized
proximal to the elbow. The patient's elbow is extended from a
position of maximal possible flexion to maximal possible
extension over a period of about 1 second by counting "one
thousand and one." In total, four passive extensions are
performed on the affected upper limb. To score the MAS the
examiner must take an average based on the four passive
extensions and disregard any volitional resistance or assistance
to movement. Any limitations to full range of passive movement at the elbow on the affected and unaffected side are
measured with a goniometer and recorded.
Two raters, both senior physiotherapists with clinical experience in neurology, underwent training, discussion, and practice
in the use of the guidelines for the TAS and MAS. Subjects were
divided into two groups (A and B), with each group containing
approximately equal numbers. Subject groups were assessed
using the TAS and MAS over 4 days in a defined counterbalanced order (table 3). Within the groups, subjects were assessed
in the same order, with attempt being made to assess an
individual at approximately the same time on consecutive days.
Within individuals, order of use of the TAS and MAS was
counterbalanced, with rater compliance being maintained by
Table 4: Interpretation of Kappa Statistic as Suggested by Brennan
and Silman 14
Kappa Statistic

Strength of Agreement

<.21
.21-.40
.41-.60
.61-.80
.81-1.00

Poor
Fair
Moderate
Good
Very good

MEASURES OF SPASTICITY, Gregson

Table 5: Agreement Between Raters for the TAS
Question
Number
1
2
3
4
5
6
7
8
9
10
11
12

Observed
Agreement
.77
.58
.74
.62
.42
.52
.37
.37
.31
.90
.41
.90

KW

.50
.22
.26
.92
.87
.87
.86
.79
.85
.79
.61
-,05

the use of previously assembled patient packs. Separate packs

were used for each assessment to negate observer bias between
and within the raters.

Statistical Analysis
Agreement between raters and within one rater was calculated for each of the 12 scale items using kappa statistic (K) for
those items which were dichotomous and kappa with quadratic
weights (Kw) for all other items) The kappa statistic is superior
to using percentage agreement alone because it corrects for
agreement by chance. 13 The results were interpreted as suggested by Brennan and Silman 14 (table 4).
RESULTS
Thirty two patients were included. The median age was 74
yrs (interquartile range, 69 to 80). There were 18 men and 14
women. The median time poststroke was 48 days (interquartile
range, 21 to 77 days). There were 27 infarcts and 5 primary
hemorrhages. OCSP classification showed 3 total anterior
circulation strokes, 15 partial anterior circulation strokes, 7
lacunar strokes, and 2 posterior circulation strokes; 5 were
unclassifiable by the information provided. There was documented history of previous stroke or transient ischemic attack in
12 cases. The median Barthel score was 8 (interquartile range, 4
to 13).
The data demonstrating agreement between and within raters
are summarized in tables 5 through 7.
Q4 to Q9 of the TAS relate to response to passive movement.
Table 5 shows that this section of the scale demonstrated good
to very good interrater reliability and table 6 shows moderate to
very good intrarater reliability, which suggests that Q4 to Q9
Table 6: Agreement Within Rater for the TAS
Question
Number

Observed
Agreement

1
2
3
4
5
6
7
8
9
10
11
12

.78
.81
.71
.58
.50
.38
.50
.63
.45
.74
.69
,76

KW

.55
.29
,39
.84
.76
.72
.59
.86
.81
.83
.78
.19

Table 7: Agreement Between and Within Raters for the MAS

Strength of
Agreement
Moderate
Poor
Poor
Very good
Very good
Very good
Very good
Good
Very good
Good
Good
Poor

Strength of
Agreement

Moderate
Fair
Fair
Very good
Good
Good
Moderate
Very good
Very good
Very good
Good
Poor

1015

Interrater
Intrarater

ObservedAgreement

KW

Strengthof agreement

.66
.32

.84
.83

Very good
Very good

may be useful to assess tone at other joints as well as at the

elbow. This is comparable with our previous findings (unpublished observations).
The questions relating to posturing at rest (Q1 through Q3)
and associated reactions (Q10 through Q12) demonstrate variability in both interrater (table 5) and intrarater (table 6)
reliability from poor to moderate. This is again similar to our
previous findings (unpublished observations). This suggests
that despite written guidelines, increased rater training, and a
larger sample size, these measures are not reliable enough to
have any clinical or research applications.
Interrater and intrarater reliability (table 7) were both very
good for the MAS at the elbow, supporting the results from
previous known studies. 4
DISCUSSION
The MAS w a s found to be a reliable measure of increased
muscle tone at the elbow, which is in agreement with other
studies. 5 The development of standardized, written guidelines
for the measurement of the MAS allows improved rater
compliance, which may in turn further improve interrater and
intrarater agreement. This is of importance because the MAS is
already being used as an outcome measure in therapeutic trials. 7
The reevaluation of the TAS showed that the parameters of
the TAS relating to posture and associated reaction were
unreliable, which is similar to our previous unpublished findings. It may be that factors other than muscle tone influence
posture, eg, balance. Associated reaction is an indirect observation and also may be better addressed as a dichotomous variable
(present or not) rather than by attempting to ascribe one point of
a four-point scale (Q10 and Qll). Alternatively, the written
guidelines for the TAS may still not be precise enough to ensure
standardization of measurement in that they may yet be open to
different interpretation by different raters.
Those questions in the TAS that relate to response to passive
movement were found to be reliable measures of muscle tone.
This is comparable to similar findings of the MAS at the elbow
and wrist, 5,8 but an advantage of this section of the TAS is that it
identifies spasticity at other joints and therefore may have wider
applicability. The question related to spasticity at the ankle
demonstrated only moderate agreement, which suggests that the
measurement of tone may truly be unreliable at this site. This
may be due to the small range of normal movement in the anne.
Also, tone is not static and may change with time, such that
repeated measurements might detect and possibly even cause
variability in tone.
In this study, reliability of both the MAS and the related
section of the TAS was established even though there were no
inclusion or exclusion criteria relating to impaired cognition or
"known" spasticity. This is in contrast to previous studies that
applied these criteria inappropriately, therefore excluding a
potentially valuable section of the available cohort. 5,8 The
findings of this study are therefore important because they
allow for wider clinical application of these scales.
CONCLUSION
The section of the TAS relating to response to passive
movement is reliable when applied to the specific joints
evaluated in this study, with the exception of the ankle.
Arch Phys Med Rehabil Vol 80, September 1999

1016

MEASURES OF SPASTICITY, Gregson

Although the MAS was reliable, the TAS may have wider
applicability because it examines tone at more joints. The TAS
may assist in future studies that seek to establish the prevalence
of poststroke spasticity and in studies of the relationship
between abnormal tone and impaired function. The TAS is not
reliable for measuring posture or associated reactions. Measures of spasticity at the ankle need to be further refined.
References
1. Wallesch C-W, Maes E, Lecomte P, Bartels C. Feasibility study on
pharmacoeconomics of botulinum toxin A (Botox) in spasticity
following stroke [abstract]. In: 3rd European Botulinum Toxin
Symposium Abstracts; 28 Feb-2 Mar 1997; London. London:
Rapid Science; 1997. p. 1351.
2. Lance JW. Control of muscle tone, reflexes and movement: Robert
Wartenberg Lecture. Neurology 1976;30:1303-13.
3. Haas BM. Measuring spasticity: a survey of current practice
among health-care professionals. Br J Ther Rehabil 1994;1:90-5.
4. Mahoney FI, Barthel DW. Functional evaluation: the Barthel
Index. Md State Med J 1965;14:61-5.
5. Bohannon RW, Smith MB. Inter-rater reliability of a modified
Ashworth scale of muscle spasticity. Phys Ther 1987;67:206-7.
6. Sloan RL, Sinclair E, Thompson S, Pentland B. Inter-rater
reliability of the modified Ashworth scale for.spasticity in hemiplegic patients. Int J Rehabil Res 1992;15:158-61.
7. Sampaio C, Ferreira JJ, Pinto AA, Crespo M, Ferro JM, CastroCal-

Arch Phys Med Rehabil Vol 80, September 1999

8.

9.

10.

11.

12.

13.
14.

das A. Botulinum toxin type A for the treatment of arm and hand
spasticity in stroke patients. Clin Rehabil 1997;11:3-7.
Bodin PG, Morris ME. Inter-rater reliability of the modified
Ashworth scale for wrist flexor spasticity following stroke. In:
Proceedings Book II, World Federation of Physical Therapy, llth
Congress; 1991; London. p. 505-7.
Taylor BA, Ellis E, Haran D. The reliability of measurement of
postural alignment to assess muscle tone change. Physiotherapy
1995;81:485-90.
Noel G. Clinical changes in muscle tone. In: Desmedt JE, editor.
New developments in electromyography and clinical neurophysiology. Basel: Karger; 1973. p. 15-9.
Dvir Z, Panturi E. Measurement of spasticity and associated
reactions in stroke patients before and after physiotherapeutic
intervention. Clin Rehabil 1993;7:15-21.
Bamford J, Sandercock R Dennis M, Warlow CA. Prospective
study of acute cerebrovascular disease in the community: the
Oxfordshire Community Stroke Project. 2. Incidence, case fatality
rates and overall outcome at one year of cerebral infarction,
primary intracerebral and subarachnoid haemorrhage. J Neurol
Neurosurg Psychiatry 1990;53:16-22.
Cohen J. A co-efficient of agreement for nominal scales. Educ
Psychol Measurement 1960;20:37-46.
Brennan P, Silman A. Statistical methods for assessing observer
variability in clinical measures. BMJ 1992;304:1491-4.