You are on page 1of 7

189-

Visual feedback after stroke with the balance


performance monitor: two single-case studies
CM Sackley and BI Baguley Division of Stroke Medicine, City Hospital, Nottingham

Research over the past decade had indicated the effectiveness of visual
feedback as a method of training stance symmetry and weight-transference
after stroke. This study was carried out to assess the efficacy of the Balance
Performance Monitor (BPM) in providing feedback. A reversal ABAB single-case
experimental design was used with two patients at different stages poststroke.
Assessments of motor function and independence in functional tasks (ADL)
were made, as well as the measures of stance symmetry. The results indicated
large improvements in symmetry, with both patients achieving levels within the
normal range after five treatments. Functional skills also improved. Although the
limitations of single-case studies are recognized, such dramatic improvements
suggest that the BPM is an effective method of providing feedback and that this
approach to treatment enhances the effects of physiotherapy and could be used
more frequently after stroke.

Introduction stand with an abnormally large amount of their


bodyweight through their unaffected leg, with
The most frequentlynoted sequela of motor estimated frequencies ranging from 61%3 to over
impairment after stroke is hemiplegia, which 80%.~ Three studies have calculated the mean

produces asymmetry of both gait and posture. percentage of bodyweight the unaffected legon
Patients tend to exhibit abnormal patterns of and estimates range from 70% to 74% of total
stance and gait; these are described extensively in bodyweight,3,7,8 far outside the normal ranges of
physiotherapy texts.l>2 Such texts also claim that between 43 % and 57% .99
patients favour their unaffected leg for weight- The clinical importance of poor stance symme-
bearing in standing and walking. try is dependent on its relation with impaired
In recent years research evidence has substan- motor function and resulting dependence in
tiated the clinical observations that patients do functional abilities. Increased weight-bearing on
the unaffected leg is directly linked with a
Address for correspondence: CM Sackley, Division of Stroke decrease in motor and functional performance.3
Medicine, City Hospital, Hucknall Road, Nottingham, UK. This relationship has been demonstrated at two

Downloaded from cre.sagepub.com at The University of Iowa Libraries on April 19, 2015
190

months after stroke and still persists at six trained an average of 30-45 minutes per day,
months. five days a week, for three to four weeks.
Conventional therapeutic intervention, using Both the experimental group and a control
neurofacilitatory techniques or concepts (such group of 21 matched subjects received conven-
as the Bobath approach), intended to
are tional physiotherapy during the trial period. At
encourage normal movement, part of which the end of the study the experimental group
involves correcting weight-bearing abnormalities. achieved a more symmetrical standing posture
Unfortunately these techniques have not been than the control group, but this difference did
formally evaluated and the theoretical ration- not generalize to measures of gait performance.
ale supporting their use is now a matter of However, the adequacy of the control group may
controversy.10 However, methods of augmented be questioned, especially as the experimental
feedback, based on motor-learning theoiy from group received more therapy treatment time in
a behavioural background have demonstrated total.
positive results. 7,11-13 A report of two single-case ’ABAB’ designs&dquo;
Since 1970 external methods of feedback have described the effects of visual feedback on weight
been used to enhance skill acquisition after distribution and gait. Patients received EMG
stroke. The advantage of using equipment to biofeedback to the arm in week A (to act as
enhance the verbal and tactile feedback from the a control) and visual feedback training in week

therapist is that it is quantifiable and precise B. An immediate and substantial improvement in


It can also provide more sensitive measures than the symmetry of weight distribution was demon-
the visual and tactile subjective evaluation of the strated, and this was sustained over time. Some
therapist. 15 carry over to the gait pattern was seen, but there
The effects of augmented feedback on postural was no measurable improvement in gross motor
control have been studied, and research using function. This study does not give the frequency
postural feedback devices has demonstrated or duration of treatment, so the total treatment

improvements in standing posture. The first time is unclear, and it may therefore be difficult
of these studies was that by Wannstedt to evaluate behavioural changes.
and Herman,16 who used augmented sensory A recent randomized controlled trial examined
feedback via an auditory signal provided from the effects of visual feedback and the transfer of
a limb-load monitor. Of the 30 ambulatory training to motor and ADL function. 17 Twenty-six
stroke patients included, 27 (77%) corrected stroke patients were randomly allocated either
their symmetry of weight-bearing by enhancing to visual feedback or to a placebo computer
their limb-loading on the paretic limb. program. Patients received 12 sessions of 20
More recently, Shumway-Cook et at.7 used minutes as part of a functionally orientated
force-plate information to provide centre of physiotherapy session (lasting approximately
pressure feedback to 16 hemiplegic patients one hour), three times a week. Independ-
and to provide an accurate assessment of stance ent assessments of motor and ADL function
symmetry. This was compared with conventional were completed before, after the four weeks’
therapy in eight control patients. The experimen- treatment and again eight weeks after that.
tal group was found to have significant increases Significant improvements in stance symmetry,
in symmetry of limb-loading. gross motor function and ADL performance were
The above studies assumed that postural observed for the treatment group at the four-week
symmetry was linked to functional ability and assessment.
that patients would automatically be able to The Balance Performance Monitor (BPM) was
transfer their newly learned skills to functional, developed to provide the therapist with an
task-orientated movement. Two studies have efficiency and effective method of using feedback
examined the transfer of skills learned from techniques (both visual and audio) in a clinical
postural training to gait, with conflicting results. environment. Unlike the equipment used in previ-
Winstein et al, 12 provided visual feedback from ous research studies it was important that the

force-place information to an experimental group BPM possessed the qualities that would enable
of 17 hemiplegic patients. These subjects were it to be used successfully within the constraints

Downloaded from cre.sagepub.com at The University of Iowa Libraries on April 19, 2015
191

of clinical rather than pure research practice. Her stroke resulted in a left hemiplegia with no
Although the results of past studies support the speech deficits but some perceptual problems.
use of augmented feedback to enhance the effects She was admitted to a general medical ward and
of rehabilitation, especially physiotherapy, it was transferred to a specialist rehabilitation unit after
necessary to test the ability of the BPM to provide five weeks; she commenced the study at seven
this facility. For this reason a detailed systematic weeks poststroke.
investigation of the effects on individual patients
(a single-case experimental design) was chosen Assessments
as it provides a large amount of detail of the
effects of treatment. Such effects may be lost The Rivermead Motor Assessment
Motor function was assessed with the
when examining a group of patients such as
Rivermead Motor Assessment.2° This scale
stroke patients, where the lack of homogeny and
was developed to provide a standardized,
diversity of symptoms make a large comparative ranked assessment of physical recovery after
sample difficult to achieve and the clinical validity stroke. The inter-rater and test-retest reliability
of group comparisons questionable. This method
have been found to be acceptable. It was designed
of investigation is becoming more popular in the
as a tool both for routine clinical use and research
field of rehabilitation research. 18
purposes. The Rivermead Motor Assessment
consists of three parts, each consisting of between
Methods 10 and 15 items arranged in order of difficulty. As
this test is a Guttman-scaled assessment2l there is
A reversal ABAB experimental design was used. no need to complete the section once a patient has
The patients were recruited while undertaking failed three consecutive items. The patient scores
their existing rehabilitation routine, all inter- the number of the last successfully completed item
vention was additional to this. Patients were
prior to the three failures.
randomly allocated to receive feedback training
in either the A or B section and received routine Ten Point Activity of Daily Living (ADL) scale
physiotherapy treatment in the other section. Functional ability was measured with the Ten
Subjects were included if they were presenting Point Activity of Daily Living (ADL) scale.22 This
with their first stroke with unilateral symptoms is a Guttman-scaled assessment and is reliable if
and were within the age range of 30-80 years. administered formally (by observing behaviour)
Both subjects were able to understand instruc- or informally (self-reporting, either as an inter-
tions and give informed consent. They exhibited view or questionnaire). It includes items such
abnormal stance symmetry (greater than 7% from as feeding, dressing and toileting. Patients are
the midline) and were excluded if they had scored as ’independent’ or ’dependent’ on each
obvious lower limb deformities (e.g. amputation, item. The test is stopped after three consecutive
joint replacement) or severe visual deficits. failures and the score (which summarizes the
patient’s level of ability) is equal to the number
Subject 1 of the most difficult item successfully completed.
KW is a male aged 42 with a history of hyper-
tension and chronic renal failure. It was his first
stroke and resulted in a left hemiplegia with no Apparatus
speech or perceptual deficits. He was admitted The Balance Performance Monitor
to a general medical ward and transferred to a The Balance Performance Monitor (BPM)
specialist rehabilitation unit after 12 weeks and (SMS Healthcare, Elizabeth House, Elizabeth
began the research study at 14 weeks poststroke. Way, Harlow, Essex CM19 5TL, UK) is a light-
At this stage poststroke it is considered that the weight portable unit designed to provide a variety
majority of recovery has already taken place.l9 of visual and auditory feedback. It provides objec-
tive measurements of left-right weight distribu-
Subject 2 tion, postural sway and anterior-posterior weight
DG is a female aged 56 with a history of transi- distribution on each foot. The equipment consists
ent ischaemic attacks over the last eight years. of the feedback unit and two movable footplates

Downloaded from cre.sagepub.com at The University of Iowa Libraries on April 19, 2015
192

(Figure 1). The software is IBM compatible and Procedure ........

can be used to provide a permanent record of the


patient’s progress. Patients were assessed while Independent assessments
Two baseline assessments, one on recruitment
standing with one foot on each of the footplates, and one after one week, consisted of the measure
which were kept a uniform distance apart, without
of stance symmetry, the balance coefficient,
any postural correction. The test lasts for 30
seconds during which the patient is asked to completion of the Rivermead Motor Assessment
stand still with their arms by their side and to (RMF) and the Ten Point Activity of Daily
fix their gaze on a suitable marker in the midline Living (ADL) scale. These assessments (as with
all others) were completed by a therapist who
(Figure 1). The BPM display was positioned away was not treating the patient and was blind to the
from the patient so the subject did not receive
any feedback during the test. The BPM calculates
patient’s treatment schedule.
Patients began with treatment A or B and were
the left-right weight distribution, described as the
assessed at the end of each week (3, 4, 5 and
’balance coefficient’, as a number from 0 to 100,
with 0 being the midline and deviation from this to 6). Subjects were then assessed at a follow-up in
weeks 7 and 8 (Figure 2).
the left or right as 0-100.
The visual feedback training was given
during week A and additional conventional
physiotherapy based on the Bobathl approach
in week B. Both treatments lasted approximately
one hour, depending on the patient’s tolerance.
Feedback from the BPM was incorporated into
a physiotherapy treatment which consisted of
a ’warm-up’ period followed by approximately

twenty minutes visual feedback, finished by


practising the new skills in a functional task,
such as gait-training. During the feedback session
patients were asked to perform a variety of tasks
aimed at improving stance symmetry. These
increased in difficulty, ranging from sitting
to standing, stride and step-standing (for a
detailed account of treatment with the BPM
see Sackley et al.23). Treatment was orientated
towards dynamic, functional skills.

Results

Subject 1 started feedback therapy in week 4 and


subject 2 in week 3. Table 1 gives the balance
coefficient data and demonstrates the improve-
ment in balance coefficient scores that occurred
during the feedback treatment (A). Figures 3 and
4 demonstrate the changes graphically. Tables 2
and 3 give the functional scores.

Discussion .

After one week’s visual-feedback treatment (five

Figure 1 The Balance Performance Monitor (BPM): sessions) both patients had reduced their stand-
feedback unit ing asymmetry and regained a normal pattern of

Downloaded from cre.sagepub.com at The University of Iowa Libraries on April 19, 2015
193

Table 2 Activity of daily living scores .

NT no visual feedback treatment


=

T = visual feedback

Table 3 Motor function scores

Figure 2 Flow chart of study procedure. BC balance


=

coefficient. RMF =
Rivermead Motor Function Assessment.
ADL Activity of daily living assessment. All intervention
=

was in addition to the patients’ rehabilitation programme


(which included physiotherapy).

Table 1 Balance coefficients


NT no visual feedback treatment
=

T = visual feedback

weight distribution. This agrees with the studies


discussed in the introduction7,11,12,16,17 in finding
that augmented feedback may improve stance
symmetry after stroke. It also demonstrates that
the BPM is an effective tool for providing visual
feedback.
Some of the improvement was lost in week
NT no visual feedback treatment
= B (additional physiotherapy without visual
T visual feedback
=
feedback), but the follow-up assessments

Downloaded from cre.sagepub.com at The University of Iowa Libraries on April 19, 2015
194

demonstrated that the effects of treatment


were retained. The improvements were gained
very quickly with very little treatment when
compared with the intense and prolonged courses
of therapy that patients receive after stroke. The
treatment schedules of 10 treatments given in
groups of five were chosen arbitrarily; it may
be more effective in the clinical environment to
start with an initial burst of treatment and then
provide regular ’top-ups’.
Transfer of training effects was seen and both
motor function and ADL scores increased after
the first week of visual feedback. Although the
numerical changes appear to be minor, the clinical
significance of these is great. The scales chosen,
especially the Ten Point ADL, are very short
and rather insensitive, so any changes recorded
are clinically relevant. The most difficult item of
the Ten Point ADL is bathing, and so a ceiling
effect can be seen when patients progress to more
difficult tasks, which may have been the case in
this study.

Figure 4 Graph of balance coefficients at each weekly


assessment for
subject 2

Single-case research designs have become


increasingly popular as a method of evaluating
physiotherapy information,18 although some
disagreement exists as to the number of data
points which should be collected. Kazdin24 recom-
mends the multiple collection of data throughout
the week, whereas Offenbacher25 considers that
patient fatigue and the limitation of the clinical
environment require that fewer assessments be
made (as in this case).
Although it is not possible to generalize the
results of two patients to the stroke population
as a whole, an interesting observation is that
visual feedback training was effective with two
patients at different stages after stroke. One
could be considered chronic (longer than three
months poststroke) and the other acute (within
three months). A larger study may indicate the
time at which it is most effective.
From the results of this and other studies
Figure 3 Graph of balance coefficients at each weekly it seems reasonable to recommend that visual
assessment for
subject 1 feedback is used more widely as an adjunct

Downloaded from cre.sagepub.com at The University of Iowa Libraries on April 19, 2015
195

to physiotherapy after stroke, especially so as 11 De Weerdt WG, Crossley SM, Lincoln NB, Harrison
there is little research evidence to support MA. Restoration of balance in stroke patients. A
other treatment methods. The results of this single case study design. Clin Rehabil
1989; 3: 139-47.
12 Winstein CJ, Gardner ER, McNeal DR, Barto PS,
study suggest that it may be a more efficient Nicholson DE. Standing balance training: effect on
method of retraining, enhancing the effects of balance and locomotion in hemiparetic adults. Arch
physiotherapy, as both patients had received daily Phys Med Rehabil
1989; 70: 755-62.
therapy since their stroke and yet had not learned 13 Crow JL, Lincoln NB, Nouri FM. The effectiveness
to weight-bear on their affected leg within the of EMG biofeedback in the treatment of arm function
normal range. after stroke. Int Disabil Stud 1990;
11: 155-60.
14 Wolf SL. EMG biofeedback applications in physical
To conclude, visual feedback with the BPM
rehabilitation: an overview. Physiother Can 1979; 31:
enhanced the effects of physiotherapy and 65-72.
improved stance symmetry; there was also a 15 Basmajian JV, De Luca CJ. Muscles alive - their
transfer of training effects to functional skills. functions revealed by EMG, fifth edition. Baltimore:
Williams & Wilkins, 1985.
16 Wannstedt GT, Herman RM. Use of augmented
References sensory feedback to achieve symmetrical standing.
Phys Ther 1978; 58: 553-59.
1 Bobath B. Adult hemiplegia: evaluation and treatment, 17 Sackley CM. A randomised controlled trial of visual
second edition. London: Heinemann, 1978. feedback after stroke. Proceedings of the 11th
2 Davies PM. Steps to follow. Berlin: Springer-Verlag, Congress of the World Confederation for Physical
1985 : 266. Therapy, London, 1991.
3 Sackley CM. The relationships between weight 18 Riddoch J, Lennon S. Evaluation of practice: the
bearing asymmetry after stroke, motor function and single case study approach. Physiother Theory Pract
activities of daily living. Physiother Theory Pract 7:
1991;
3-11.
1990;
6:179-85. 19 Wade DT, Langton Hewer R, Skilbeck CE, David
4 Dickstein R, Nissan M, Pillar T, Scheer D. RM. Stroke, a critical approach to diagnosis, treatment
Footground pressure patterns of standing hemiplegic and management. London: Chapman & Hall, 1985:
patients. Phys Ther 1984; 64:
19-23. 52-53.
5 Caldwell C, MacDonald D, Macneil K, McFarland K, 20 Lincoln NB, Leadbitter D. Assessment of motor
Turnbull GI, Wall JC. Symmetry of weight distribu- function in stroke patients. Physiotherapy 1979; 65:
tion in normals and stroke patients using digital weigh 48-51.
scales. Physiother Pract 1986; 2: 109-16. 21 Williams HGA, Johnston M, Willis LA, Bennett AE.
6 Bohannon RW, Larkin PA. Lower extremity Disability: a model and measurement technique. Br J
weight bearing under various standing conditions in Prevent Soc Med 1976; 30: 71-78.
independently ambulatory patients with hemiparesis. 22 Ebrahim S, Nouri F, Barer D. Measuring disability
Phys Ther 1985; 65:
1323-25. after a stroke. J Epidemiol Community Health 1985;
7 Shumway-Cook A, Anson D, Haller S. Postural 39:
86-89.
sway biofeedback: its effect on re-establishing stance 23 Sackley CM, Baguley BI, Gent S, Hodgson P.
stability in hemiplegic patients. Arch Phys Med Use of the balance performance monitor in the
1988; 69: 395-400.
Rehabil treatment of weight-bearing and weight-transference
8 Mizrahi J, Solzi P, Ring H, Nisell R. Postural stability problems after stroke. Physiotherapy 1992; 72:
in stroke patients: vectorial expression of asymmetry, 907-913.
sway activity and relative sequence of reactive forces. 24 Kazdin AE. Single-case research designs. Methods
Med Biol Eng Comput 1989; 27: 181-90. for clinical and applied settings. Oxford: Oxford
9 Sackley CM, Lincoln NB. Weight distribution and University Press, 1982: 104.
postural sway in healthy adults. Clin Rehabil
1991; 5: 25 Offenbacher KJ. Evaluating clinical change. Balti-
181-86. more: Williams & Wilkins, 1986: 43-48.
10 Bohannon RW. [Correspondence]. Physiother Theory
Pract 1990; 6: 246.

Downloaded from cre.sagepub.com at The University of Iowa Libraries on April 19, 2015

You might also like