Professional Documents
Culture Documents
Submitted DMCNArticle Knox Evans May 2001
Submitted DMCNArticle Knox Evans May 2001
Abstract
This study aimed to evaluate the functional effects of Bobath therapy in
children with cerebral palsy. Fifteen subjects, aged 2-12 years, all having a
diagnosis of cerebral palsy, were recruited. Children having orthopaedic
intervention planned were excluded. A repeated measures design was used
with subjects tested with the Gross Motor Function Measure (GMFM) and
Pediatric Evaluation of Disability Inventory (PEDI) at six weekly intervals
(baseline, before and after Bobath therapy, and follow-up). As the data was of
ordinal type, non parametric statistics were used, i.e. Wilcoxon Test. Subjects
showed a significant improvement in scores in the following areas following
Bobath treatment compared to the periods before and after the Bobath
therapy: GMFM total score (P=0.009); GMFM Goal Total (p=0.001); PEDI self
care skills (P=0.036); and PEDI caregiver assistance (P=0.012). This
demonstrates that in this population, following a block of Bobath therapy,
gains were made in motor function and self care.
Introduction
At present it is difficult to demonstrate the effects of physiotherapy on children
with cerebral palsy. This is because of the diversity in all aspects: the different
motor disorders, the different treatment approaches which lack clear
description and the lack of suitable validated evaluative tools. Hur (1995)
reviewed 37 studies of therapeutic interventions for children with cerebral
palsy, and reported that the majority of the studies had small samples, were
poorly controlled and some had a lack of rigour in both experimental design
and analysis. Of the seven groups using a comparative design, only two
showed a significant treatment effect.
One therapy approach, most widely used within the UK for children with
2
cerebral palsy is the Bobath approach (Bobath & Bobath, 1984). The Bobath
concept emphasises observation and analysis of the client’s current functional
skill performance and the identification of clear therapy goals. The aim of
treatment is to influence muscle tone and improve postural alignment by
specific handling techniques, and then to work for better active participation
and practise of specific relevant functional skills. It is considered to be
appropriate for treating any motor control disorder within the cerebral palsy
spectrum. (Mayston,1992; Mayston et al, 1997). Treatment programmes
within the Bobath concept are goal focussed. It is forward looking, focusing on
the likely potential for secondary deformities and how these may be
prevented. Parent/carer education is one of the main elements of the
intervention: to facilitate the parent child relationship, to enable the parent to
handle/assist with their child’s difficulties appropriately, and give an intensive
period for practise of activities (Mayston, 1992; Bly, 1991).
There has been a lack of rigorous research into the clinical effectiveness of
Bobath (neurodevelopmental - NDT) therapy (Royeen & DeGangi, 1992).
Ottenbacher et al (1986) conducted a meta analysis of studies which had
investigated the effects of NDT in paediatric populations. This showed that
clients receiving NDT or a combination of NDT and other intervention
performed better than 62% of subjects receiving other treatment modalities,
although the effect size was perceived to be small. Some other small studies
have also shown NDT may be of benefit (Jonsdottir et al, 1997; Kluzik, 1990;
De Gangi, 1994; Laskas, 1985). However, Royeen and DeGangi (1992)
reviewed 19 studies investigating the effects of NDT, finding many to have
inconclusive results. There were similar problems with sample size, lack of
suitable validated measures and experimental design, as for research into all
therapy for children with cerebral palsy.More studies investigating the efficacy
of specific interventions are needed which use appropriate experimental
designs (Hur, 1995; Royeen & De Gangi, 1992). The purpose of the present
study was to investigate the effects of a six week block of Bobath therapy on
the function of children with cerebral palsy, this being current practice in
Centres run according to the Bobath concept.
Method
Experimental Design
A repeated measures design was used with assessment carried out at six
weekly intervals: at baseline, prior to Bobath treatment, after Bobath treatment
and at follow up. In this pilot study, the children acted as their own controls.
The amount of local therapy received by subjects was not altered, but was
recorded by the parents.
Participants
Twenty subjects (aged 2 - 12 years, boys n=12; girls n=8) who were referred
3
Materials
Two standardised validated measures of function were used:
(A) The Gross Motor Function Measure (GMFM) which assesses gross motor
abilities of children with cerebral palsy, in five dimensions: (1) lie and roll, (2)
sit, (3) crawl and kneel, (4) stand, and (5) walk, run and jump (Russell et al.
1989, 1993). In children with cerebral palsy, the GMFM has been shown to be
sensitive to change during periods of therapy (Bower et al. 1992, 1994a,
1996; Steinbok et al, 1997).
(B) The Paediatric Evaluation of Disability Inventory (PEDI) which assesses
mobility, selfcare and social function across three scales: functional skill level,
physical assistance typically required of the care giver and any modifications
or equipment used (Haley et al.1992). The Inventory is completed on parent
interview. The PEDI has been shown to be sensitive to differences between
children with differing distributions of cerebral palsy and to change following
surgery and therapy (Bloom & Nazar, 1994; Dudgeon et al.1994).
Procedure
Informed written consent was obtained from the parents. Subjects received
their baseline assessment (GMFM & PEDI) and parents were asked to record
local therapy over the course of the trial. Another assessment took place at
six weeks and a parent questionnaire was administered asking if any changes
were perceived to have taken place in the child’s mobility, activities of daily
living, play or communication. Prior to Bobath intervention, three short term
treatment goals were agreed with the family and recorded. Intervention began
and attendance for treatment sessions was recorded (expected attendance
being 16 sessions). At 12 weeks, (the end of the intervention period), the third
assessment took place and questionnaires were administered to the parents
and treating therapist regarding perceived changes in function. A final
assessment took place at 18 weeks.
Data Analysis
In addition to individual dimension scores for the GMFM, the following scores
were analysed: GMFM total, GMFM goal total1, GMFM non-goal total2 and
GMFM-66 scores. The GMFM-66 is a new method of scoring using only 66
test items that have been arranged in order of item difficulty (Russell et al,
2000). This allows an interval score to be calculated representative of the
overall level of motor ability of the child, unlike the other GMFM scores, which
represent how many and to what extent items are achieved. In addition to
individual PEDI domain scores, total scores for functional skills and caregiver
1
Goal total: determined by observing in which dimension(s) the treating Bobath therapist had set goals at onset of therapy and
assigning those dimension(s) as the Goal Total score (if more than one dimension is highlighted, an average score is calculated).
2
Non-Goal total: an average score calculated from all dimensions not assigned as goal areas.
4
Results
Of the twenty subjects recruited, five could not be followed up due to: illness
during trial (n=2); botulinum toxin injections during trial (n=1); failure to attend
majority of treatment sessions (n=1); and communication difficulties causing
the subject distress during measurement procedures (n=1). Of the 15
remaining subjects (boys n=9; girls n=6), ages ranged from 2 years 8 months
to 12 years. Types of cerebral palsy were as follows:
Treatment attendance was good for the majority of the subjects: five attended
3
Calculation of total scores are not suggested in the manual, but as this test was thought to be less sensitive over a period of a
few weeks, it was thought this might maximise the chance of seeing any change in scores that occurred.
5
for all 16 sessions; nine attended for between 11 and 15 sessions, and one
subject only attended for six sessions. Eight parents completed the local
therapy forms. Local therapy varied in frequency from none to twice weekly.
During the Bobath intervention, three reported that local therapy remained
consistent, one reported slightly less and four less regular or no local therapy.
Fifty-seven GMFMs and 53 PEDIs were completed (out of a possible 60
each). Complete data was available for nine subjects and partial sets for the
remaining six subjects, but this was sufficient for statistical analysis. Missing
data was largely from the final test at 18 weeks, six weeks after Bobath
therapy had finished, where some subjects failed to return for final testing or
parental time was limited for completion of the PEDI.
A large proportion of the goals set by the therapist were within areas covered
by the GMFM or PEDI, such as relating to sitting, self care, etc. Other goals
did not fall within the remit of either measure, for example ' Fill in a missing
part on pre-drawn picture'; 'produce breath for group of 3/4 syllables', etc, so
neither standardised test could be expected to demonstrate whether these
were achieved.
This is also illustrated in Figure 1, which presents inter-test differences for the
total scores, expressed as median differences and 25 th and 75th percentiles of
the distribution.
No significant treatment effects were seen in the scores for Sitting (p=0.259)
or Standing (p=0.507) dimensions. The subjects did show an overall
improvement in score in these areas from the beginning of the study to the
end, but this appeared unrelated to the Bobath therapy. This can be seen in
the graph displaying mean values for GMFM scores (see Figure 2).
Due to the heterogeneity of this group of subjects consistent with the variety
of function ability seen within children with cerebral palsy, the standard
deviations are very large. For example, some children were unable to achieve
a score on any items within the walking dimension whereas other children
achieved a score of up to 88% (see Table 4). The standard deviations are
7
Discussion
In this small pilot study, subjects showed a significant improvement over the
six week intervention period, compared to the pre and post treatment scores,
in the following areas of gross motor function: walking (p=0.010); crawling
(p=0.05); GMFM total (p=0.009) and Goal Total scores (p=0.001). The score
for lying nearly reached significance (p=0.66).
The most significant result was seen in the goal totals (p=0.001). No
significant change was seen in the non-goal total score (p=0.196), which
suggests that change was achieved in the goal areas where therapy was
concentrated. This provides some support for the study by Bower et al (1996)
where the use of goals enhanced treatment outcome over a treatment period
of 2 weeks. It differs from the pilot study by Bower et al (1992) where the
differences in scores between the goal and non goal totals was less apparent.
However in both this present study and that of Bower (1992), small subject
populations were used and inevitably there will be considerable diversity in
their population characteristics. Also, there was a difference in the time
periods of treatment (two versus six weeks). It differs from the randomised
controlled trial with a six month intervention period by Bower et al (2001),
where minimal benefit was observed from the use of goals and more intensive
daily treatment.
Change was not detected in the sitting dimension, although this area was
mentioned as one of the goal areas for nine subjects. Positive changes were
reported after therapy by all therapists and all but two parents. Several
reasons could account for no changes being recorded on standardised
testing.
First it is certainly possible that no overall change took place in this area
within this population of subjects, despite the reports of therapists and parents
to the contrary. They may have had bias due to having expectations, after
setting goals. Secondly, sitting was the dimension where the largest number
of subjects started with a score of >95 leaving a smaller degree of freedom for
change (n=8 at baseline), demonstrating the ‘ceiling effect’ of the GMFM.
Thirdly, some subjects were reported to have an improvement in a skill or
quality of sitting not tested by the GMFM, such as reaching while bench
sitting. In conclusion, there may have been a Type II error, that is a change
occurred in sitting function, which was not detected by the standardised
measures used in this study.
Change was not detected in the standing dimension, with seven subjects
having this as a goal area. There was little difference between the test results
of those that had standing as a goal area and those subjects that did not.
10
Improvement was not seen overall in the mobility skill section of the PEDI.
Ten subjects had goals set within this domain. Although the skill level of
subjects as measured on the PEDI did not change during the treatment
phase, the level of caregiver assistance required for such tasks decreased.
Improvements were not seen in social function. Only a small proportion of this
group of children were either seen by a speech and language therapist during
their therapy or had goals set in this area (n=3). The Social Function domain
provides only a limited assessment of communication skills with quite large
steps between each item/level and was not sufficiently sensitive to detect
changes likely to occur over a six week period.
Although the two standardised measures chosen are among the best
available for children with cerebral palsy (Ketelaar, 1998), they do have
limitations. It was apparent from the parent and therapist questionnaires that
other changes were perceived to occur, both in improved functional skills and
quality of movement. The GMFM and PEDI only measure certain aspects of
function and do not purport to measure how a child performs a task such as
speed, co-ordination and fluency of movement (Wright et al, 1998).
To obtain a maximum score on the GMFM, the child must attempt as many
items as possible. However, if a child can achieve an item at a higher level,
such as crawling, they may be reluctant to attempt an item at a lower level,
such as creeping in prone. Children can therefore possibly function at a higher
level, but achieve a lower or similar score due to refusal or poor attempts at
lower level items. Subject 18 was very reluctant to attempt the easier items
particularly those on the floor, as he could function very well when up against
gravity in standing/walking.
Floor and ceiling effects can affect the sensitivity of a test. If a child is already
11
performing at near 100% or baseline, there are limited degrees of freedom for
change (Russell et al, 2000; Stanley et al, 2000). McClaughlin et al (1998)
suggested that there might be a difference in level of difficulty of items in
different parts of the range of the GMFM scores with the upper range being
less sensitive to change. The scale is ordinal and differences between scores
are not intended to represent equal differences of ability.
No control group was studied, the children acting as their own controls in this
study; this together with the small sample size does limit the power of the
results. For the planned main trial, subjects will be stratified into age bands
and GMFCS levels, and randomised into one of two groups: a treatment
group as described in this study; and a control group which will be measured
at the same time intervals with no Bobath therapy being instigated.
Conclusion
Within this population sample, subjects improved significantly in gross motor
and self-care skills and required a lower level of caregiver assistance for
mobility and
self care, as measured on the GMFM and PEDI, following a 6 week block of
Bobath therapy. Most improvements occurred within areas in which therapy
goals were set. If this can be confirmed by a larger randomised controlled
trial, then it may provide useful information to empower parents, therapists
and purchasers in decision making regarding appropriate therapy.
References
12
Tardieu C, Lespargot A, Tabary C, Bret M. (1988) For how long must the
soleus muscle be stretched each day to prevent contracture?
Developmental Medicine and Child Neurology, 30: 3-10
Watt J. et al (1986) A Prospective Study of Inhibitive Casting as an Adjunct to
Physiotherapy for Cerebral Palsied Children. Developmental Medicine
and Child Neurology, 28:480-488
Wright V, Sheil E, Drake J, Wedge J, Naumann S. (1998) Evaluation of
selective dorsal rhizotomy for the reduction of spasticity in cerebral
palsy: a randomized controlled trial. Developmental Medicine and Child
Neurology, 40:239-247