Professional Documents
Culture Documents
Abstract
Introduction. This replicated randomized controlled crossover case series investigated the effect of mirror-based
tactile and motor training on tactile registration and perception in children with unilateral cerebral palsy (UCP).
Methods. Six children with UCP (6–18 years; median 10 years, five male, three-left hemiplegia, four-manual ability
classification system (MACS) I, one MACS II and one MACS III) participated. They attended two 90-minute ses-
sions — one of mirror-based training and one of standard practice, bimanual therapy — in alternated order. Tactile
registration (Semmes Weinstein Monofilaments) and perception (double simultaneous or single-point localization)
were assessed before and after each session. Change was estimated using reliable change index (RCI). Results. Tac-
tile perception improved in four participants (RCI > 1.75), with mirror-based training, but was unchanged with bi-
manual therapy (RCI < 1.0 for all participants). Neither intervention affected tactile registration. Discussion.
Mirror-based training demonstrates potential to improve tactile perception in children with UCP. Copyright ©
2016 John Wiley & Sons, Ltd.
Keywords
cerebral palsy; mirror therapy; tactile; upper limb
*Correspondence
Megan Auld, PhD, School of Health and Rehabilitation Sciences, The University of Queensland, Brisbane, QLD 4072, Australia.
E-mail: mauld@uq.edu.au
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
Mirror-based Training in Cerebral Palsy M. Auld et al.
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
M. Auld et al. Mirror-based Training in Cerebral Palsy
score >10; double-simultaneous score >21), the child 20-seconds inter-stimulus interval) on their im-
was excluded without receiving unnecessary treatment, paired hand inside the mirror box, randomly given
and their data were not analysed. to one of five locations. During stimulation, partici-
Tactile registration Tactile registration was measured pants looked in the direction of their impaired hand
using the full 20-item Semmes–Weinstein monofila- but in doing so saw the image of their unimpaired
ment kit (Bell-Krotoski, 1990). hand in the mirror. After each stimulus, participants
Spatial tactile perception Single-point localization was responded by pointing to the location on a picture
assessed using the largest of the monofilaments (from of a hand or verbally indicating the finger that was
the Semmes–Weinstein Monofilaments) according to touched. There were 24 stimuli in a block, which
Auld et al. (2012b). Double simultaneous was assessed took 8 minutes. There were three blocks, separated
using two identical tactile stimulators according to by a 3-minute rest. To maintain the child’s attention
Auld et al. (2012b). (i.e. younger children), and motivation, a game was
The beneficial effect of multisensory training should played during the test session with the child
be most obvious in measures of perception rather than progressing a game piece following each response.
registration (Gallace and Spence, 2008). As such, a hi- Some example games were the following: lego as-
erarchy of tactile perception was used, such that the sembly (every correct answer achieves another piece
primary outcome for each participant was the measure with the Lego item being constructed at the end of
for which a reliable score could be determined, in the the session) and making puzzles (every correct an-
following order: double simultaneous, single-point lo- swer the child added a puzzle piece with the aim that
calization. That is, double simultaneous was used un- they could complete the puzzle at the end of the ses-
less a reliable score could not be obtained, in which sion). So that the child did not need to move, the ex-
case single-point localization was used. To minimize aminer moved the piece of the game within the
the likelihood of false-positive findings within partici- child’s visual display, before the child turned their
pants, only one outcome was analysed. Secondary attention back to the mirror image for the next
outcome was tactile registration. Assessment was stimulus.
undertaken before and after each session. • Mirror-based movement training: In the same posi-
tion as for tactile training, participants copied
Intervention upper-limb movement positions on verbal command
as provided by the investigator. The methodology has
There were two interventions: mirror-based tactile and
been modified from Dohle et al., 2009 to be suitable
motor training, herein called mirror training, and stan-
for children. Hand positions copied were age-
dard practice bimanual motor-learning approach,
appropriate and are those specified in the upper-
herein called bimanual training.
limb motor planning examination of the Neurosen-
sory Motor Developmental Assessment (Burns,
The active treatment: mirror training
2014). No tactile stimuli were provided during this
Children sat in a quiet, well-lit room. The impaired condition. Children were asked to respond to the in-
arm was placed inside a commercially available mirror structions of the next position as quickly as possible.
box (noigroup.com, Adelaide, Australia), with the el- This section lasted up to 30 minutes per session (in
bow and wrist extended and forearm supinated. The two 15-minute blocks), moving through the stan-
unimpaired limb also rested in the extended, supinated dardized movement positions within the physical
position, aligned so that the participant saw the mirror and motor planning ability of the child.
image of the unimpaired limb as though it was the im-
paired limb. There were two components to mirror
Tactile and motor components were alternated
training.
throughout the treatment session (Table 1).
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
Mirror-based Training in Cerebral Palsy M. Auld et al.
planning abilities using repetitive practice of bimanual that we can be 90% confident, and an RCI of 1.96
activities was carried out. To ensure that tasks were means that we can be 95% confident, that there really
age-appropriate, directed play tasks similar to those was a clinically meaningful change in status for that
included in the Assisting Hand Assessment individual.
(Krumlinde-Sundholm et al., 2007) were carried out
for the treatment time (approximately 50 minutes), in Results
a manner similar to that described in the comparative
Ten children agreed to participate. Three children were
treatment in Hoare et al. (2013). Bimanual therapy
excluded. One child dropped out prior to the second
is currently considered standard care for the upper
appointment. Thus, six children with UCP between
limb in children with UCP (Sakzewski, 2012; Hoare
the ages of 6–18 years (median 10 years, five male,
et al., 2013).
three-left hemiplegia, four MACS I, one MACS II and
one MACS III) (Table 2), participated. None of the
Procedure participants received any other treatments that might
Participants attended two 1.5-hour treatment sessions have influenced tactile function during the course of
at a clinic venue or at the child’s home, carried out the trial.
by an experienced physiotherapist. At both sessions, Participant one (P1): A 7-year-old male with right-
children were assessed, treated and then assessed again. sided UCP; MACS I. Although P1 had a diagnosis of
In one treatment session, the child received mirror- right UCP throughout the assessment and treatment
based training, and in the other, the child received sessions, he regularly used two hands together and
control bimanual therapy, with the order of these would occasionally use his right hand to perform the
therapy types randomized between children. Appoint- dominant role in a task (e.g. hold a cup). At the time
ments were scheduled so that the duration between of assessment, P1 had also commenced a trial of Ritalin
sessions varied across participants between 2 and to aid concentration at school. On initial examination,
7 days, so as to comply with recommended best prac- P1 was unable to reliably report for double simulta-
tice to control for time in replicated case series designs neous and had a significant impairment in single-
(Ferron and Onghena, 1996; Onghena and Edgington, point localization and a mild impairment in tactile
2005). registration. P1 required several breaks during mirror
therapy to maintain attention.
P2: A 10-year-old female with right UCP, MACS III.
Data analysis
P2 was born at 40 + 5 weeks gestation (5 days overdue).
Reliable change index (RCI) (Jacobsen and Truax, P2 was diagnosed with right-sided UCP at 6 months of
1991) was calculated for all participants for both age. At 10 months of age, P2 began to experience
interventions. RCI provides an indication of how sure seizures, which continued until 5 years of age when
we can be that a change in status was achieved, when P2 had a hemispherectomy to alleviate these seizures.
the inherent reliability of the measurement tool An MRI at 11 months and the surgical intervention at
and the minimum clinically detectable change are 5 years confirmed a diagnosis of left-hemispheric
considered. RCI is recommended when individual polymicrogyria. Although all assessment items were
responses are analysed, or for studies that involve attempted, P2’s reporting was most consistent and
small samples, such as in a replicated case series reliable on single-point localization, in which under-
such as this. RCI is expressed as standard deviation standing of the task could be easily confirmed with
unit, such that an RCI of greater than 1.7 means her achieving perfect perception on her unimpaired
Time → (minutes) 20 8 15 8 15 8 20
Total time = 94 minutes
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
M. Auld et al. Mirror-based Training in Cerebral Palsy
Table 3. Tactile scores pre-bimanual and post-bimanual training (control) on the impaired hand
Child Pre Post Pre Post Pre Post Reliable change index
1 3 3 11 10 — — 0.56
2 — — 0 0 — — 0.00
3 4 5 11 11 21 24 0.97
4 5 5 1 2 10 10 0.00
5 3 3 12 12 19 17 0.50
6 4 5 12 10 17 17 0.00
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
Mirror-based Training in Cerebral Palsy M. Auld et al.
probably means there was little room for improvement of the variance in bimanual performance can be
anyway. Bimanual therapy, the control condition, did explained by outcomes in double simultaneous (Auld
not improve tactile perception in any participant. et al., 2012c). One might predict that improved tactile
Given that it is known that over 77% of children with perception could facilitate improved motor function,
UCP have an impairment in tactile function (Auld although clearly this needs to be empirically evaluated.
et al., 2012a) and that there are currently no known No significant changes in tactile perception were
treatments for managing this impairment (Auld noted following the one-control session of bimanual
et al., 2014), these results raise the possibility of better therapy, concurring with the previous systematic
outcomes for children with UCP. review indicating that treatments aimed at motor
Performance in single-point localization and double impairments do not also improve tactile impairments
simultaneous has an established relationship with (Auld et al., 2014). A recent study indicates that
upper-limb motor function. Indeed, over 30% of the 90 hours of bimanual therapy delivered over 3 weeks
variance in unimanual capacity can be explained by leads to improvements in tactile perception, as mea-
outcomes in single-point localization, and over 30% sured by the grating orientation task (Kuo et al.,
Figure 1. Improvement in spatial tactile perception following mirror-based training and bimanual training
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
M. Auld et al. Mirror-based Training in Cerebral Palsy
Table 4. Tactile scores pre-mirror and post-mirror training on the impaired hand
Reliable change
Registration Single-point localization Double simultaneous index
1 5 5 7 12* — — 2.78
2 — — 1 5* — — 2.22
3 5 4 12 12 24 23 0.25
4 5 5 1 2 9 16* 1.76
5 3 2 12 12 17 24* 1.76
6 4 4 11 11 16 20* 1.01
2016). This same treatment only led to a trend in initially showed the greatest improvement. Critically,
improvement in stereognosis and no improvements however, those same participants did not respond to bi-
in either spatial tactile perception as measured by manual therapy, which rules out the possibility that the
two-point discrimination or tactile registration as mea- effect of mirror training was simply a reflection of tactile
sured by the Semmes Weinstein Monofilaments (Kuo impairment severity.
et al., 2016). Notably, however, the dose in the Kuo That non-informative vision augments tactile
et al. paper was extremely high in comparison with evoked cortical responses in healthy volunteers
the current study (90 hours compared with less than (Taylor-Clarke et al., 2002) –— responses that are con-
1 hour). That clinically important change was observed sistent with primary sensory cortex activation — raised
after a single session of mirror training, strongly sug- the possibility that we would see improved
gests that it may be a potent treatment approach. There performance on tests of registration. That we did not
are data from adults with pathological pain and tactile is consistent with a recent systematic review of behav-
dysfunction that show similar improvements with ioural evidence of tactile improvement with non-
one-training session (Moseley and Wiech, 2009), and informative vision (Eades, 2015), which shows clear
in that group, repeated training has been shown to have benefits on tests of tactile perception rather than
a greater effect that is maintained for at least 3 months registration. This finding might be predicted on the
post-training (Moseley et al., 2008). grounds that multisensory processing is important for
For mirror-based training to be successful, it is awareness and occurs after the primary touch signal
essential that the child attends to the visual display of has been processed in the primary sensory cortex, yet
their unimpaired hand in the mirror for both the before a percept is generated (Gallace and Spence,
sensory and motor exercises (Kennett et al., 2001b; 2008). Further research incorporating evaluation of
Dohle et al., 2009). In the current study, Participants the effect of mirror-based training on cortically evoked
4, 5 and 6 showed the largest improvement in double- responses to touch would cast light on this issue.
simultaneous perception. They were also the eldest of Hand position during assessment and treatment may
the sample and able to complete the tasks without the also be critical to the success of mirror-based training.
interruption of games or breaks. Although attempts This was particularly relevant for Participant 5, who
were made to maintain engagement in the younger chil- achieved significant gains on double simultaneous
dren, it seems reasonable that the ability to attend was a (identifying the digits that were stimulated on both
limiting factor in improvement. One way to mitigate hands) but not on single-point localization (identify-
this may be to undertake repeated sessions that are ing the exact location of touch on the impaired
shorter, or to incorporate mirror training into a regular hand). Participant 5 held an atypical posture with
aspect of life. Another factor that probably affects re- his impaired hand throughout treatment and assess-
sponse to mirror training is the severity of initial tactile ment, so, although the therapy may have assisted his
impairment — those with the poorest performance ability to distinguish between fingers, he remained
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
Mirror-based Training in Cerebral Palsy M. Auld et al.
unable to distinguish specific locations on his age, attention and lesion characteristics to aid optimal
impaired hand. patient selection. Subsequent studies should also inves-
It is not possible to determine from the current tigate the specific treatment benefit of the separate
study whether the sensory or motor component of tactile and motor components of the mirror-based
training had more effect on tactile perception, or if training.
both were equally useful for improving function. In a
previous pilot study in children with UCP, a version Implications for physiotherapy
of motor mirror training was shown to improve practice
upper-limb strength and function as measured by the
Shriners Hospital Upper Extremity Evaluation (Gygax • A single session of mirror-based training shows
et al., 2011). However, our objective was to determine potential to improve tactile perception in children
if a broad visual-enhanced tactile and motor pro- with UCP.
gramme was beneficial, rather than attempt to untangle
the contributions of different components. If the cur-
rent results are replicated in a randomized controlled Acknowledgements
trial, such a study would clearly be warranted. The authors would like to thank the Cerebral Palsy
Interpretation of the current results should consider League for assistance with recruitment and the children
several limitations. A small sample increases the likeli- and families who participated in the study. G. L. M. was
hood of erroneous results. However, we employed the supported by a Research Fellowship from the National
gold standard design for replicated case series Health & Medical Research Council of Australia (ID
approach, which controls for time and order of condi- 1061279).
tions (Ferron and Onghena, 1996; Onghena and
Edgington, 2005). We also implemented the RCI
REFERENCES
(Jacobsen and Truax, 1991), which provides a conser-
vative estimate of how likely it is, in light of the Australian Cerebral Palsy Register (ACPR). Report of the
variability of the score in the studied cohort, the Australian cerebral palsy register birth years 2009;
reliability of the assessments used and the clinically 1993–2003.
important effect, that a true change in status occurred. Auld ML, Boyd RN, Moseley GL, Ware RS, Johnston LM.
That we can be over 90% confident that four out of six Tactile function in children with unilateral cerebral
palsy compared to typically developing children.
participants showed a clinically meaningful improve-
Disability and Rehabilitation 2012a; 34(17): 1488–1494.
ment in tactile function is very encouraging indeed.
Auld ML, Boyd RN, Moseley GL, Johnston LM. Tactile
That our sample was heterogeneous adds to the weight
assessment in children with cerebral palsy: a clinimetric
of our finding but also reduces the likelihood of finding review. Physical and Occupational Therapy in Pediatrics
effects in either condition. We did not obtain neuroim- 2011; 31(4): 413–439.
aging data for the studied cohort. This is relevant Auld ML, Ware RS, Boyd RN, Moseley GL, Johnston LM.
because the location, timing and extent of lesion may Reproducibility of tactile assessments for children with
be critical for patient selection (Staudt, 2010) and unilateral cerebral palsy. Physical and Occupational
may also influence response to mirror-based training. Therapy in Pediatrics 2012b; 32(2): 151–166.
Auld ML, Boyd RN, Moseley GL, Ware RS, Johnston LM.
Impact of tactile dysfunction on upper limb motor
Conclusion function in children with unilateral cerebral palsy.
Archives of Physical Medicine and Rehabilitation
This study demonstrated for the first time the potential
2012c; 93: 696–702.
for a 1-hour mirror-based tactile and motor training
Auld ML, Russo R, Moseley GL, Johnston LM. Determina-
session to improve tactile perception in children with
tion of interventions for upper extremity tactile impair-
UCP. That four out of six participants had a clinically ment in children with cerebral palsy: a systematic
important improvement in tactile impairment is review. Developmental Medicine and Child Neurology
promising but needs to be verified in a randomized 2014; 56(9): 815–832.
controlled trial. Such a study might also permit inves- Bell-Krotoski J. In: Hunter J, Schneider L, Mackin E,
tigation of dose response and evaluate the influence of Callahan A (eds), Light Touch — Deep Pressure Testing
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.
M. Auld et al. Mirror-based Training in Cerebral Palsy
Physiother. Res. Int. (2016) © 2016 John Wiley & Sons, Ltd.