(I was asked how can it get more complex as it "gradually" works
up to
the 3 joints of the ankle, knee and hip?)
A system of multi-linked bodies is a system - so it really doesn't
matter where one starts, as the complexity is in the system not in
any
one part of it. However, I just find it personally more insightful
and
rewarding to start at the foot end. I think this is because most of
the
pathologies we see (e.g. cerebral palsy, stroke, amputees) tend to
predominantly affect the distal joints. In fact, one of the
questions I
put on the page - the classification of hemiplegia - is based
on the
most proximal joint involved - I'd say this little girl is grade 1
(dynamic ankle spasticity only) but I'd be interested to see what others
think.
The other reason why I personally like to start at the foot-ankle is
because it provides the major power burst for propelling the leg
into
swing (and thus achieving an adequate stride length). Other joints
can
compensate, but I think it's important to look for inadequate push-off
first. This will often explain what is happening proximally.
As I say. I'd be interested to hear what others think. I'll be in China
till Monday, but I'm looking forward to rejoining the discussion then.
Chris
--
Dr. Chris Kirtley
Dept. of Rehabilitation Sciences
The Hong Kong Polytechnic University
I hope you don´t mind my comments because I am a beginner on Gait
Analysis, I am in charge of a gait Laboratory on a brand new Pediatric
Rehabilitation Center in Mexico City called Telethon, as far we have
a
waiting list of 3,000 children and I think that at least 40% wil be
seen
at the Gait Laboratory.
My comments are that the main Kinematic anormalities are at the
hip and
rigth knee, because during initial and mid stance there is a
deviation
in varus of the knee and in add, this could be due to overactivity
of
hip adductors and weakness of hip abductors. The Kinetic abnormalities
are in the right hip because during stance there is a double bump
on
power generation for hip extensors and power generation for
knee
extensor to increase knee stability, because of the plantarflexor moment
of the ankle at initial contact and initial and midstance.
The abnormality on right gastrocnemius is during stance its activity
is
low and weak to control tibial advancement and refrenate it.
I hope you don´t find my comments too basic, I am trying hard
to be a
good gait analyzer.
Best Wishes
BTW I do agree with the treatment chioce of botox
injections to her right gastroc, but where is the endpoint
of treatment?
Dev Med and Child Neurol, 39(S75):3, 1997
Kinematic and Kinetic Determinants of Stride Length in
Children with Cerebral Palsy.
MICHAEL ORENDURFF, MS; ROSEMARY PIERCE, LPT; ROBIN
DOROCIAK, BS; MICHAEL AIONA, MD. (SHRINERS HOSPITAL FOR
CHILDREN--PORTLAND UNIT, 3101 SW SAM JACKSON PARK ROAD,
PORTLAND, OR 97201, USA)
Objectives: It has been assumed that increased knee
flexion is a cause of decreased stride length in children
with cerebral palsy (CP). However, successful surgical
correction of crouch sometimes does not lead to stride
length improvements. We hypothesized that limitations to
stride length could be caused by decreased hip or knee
motion, decreased ankle push-off energy, or contralateral
stance limitations (i.e. increased double support). Leg
length was also included as it was assumed to have major
influence on stride length. Our objectives were to
determine which kinematic (motion) and kinetic (force) gait
information would predict stride length in children with CP.
Design: A prospective gait study of 54 pediatric patients
with cerebral palsy followed at an regional children's
hospital.
Participants: Data from 54 consecutive individuals with CP
who received computerized gait analysis as part of their
care were analyzed (54 sides chosen randomly from barefoot
free walk trials). Mean age was 11.8 ± 4.2 years, range
4.7 to 20.5 years. 15 normal children were used for
comparison (mean age 8.5 ± 3.1 years, range 5.8 to 14.3
years).
Methods: Multiple linear regression was used to evaluate
the effects of hip and knee kinematics, ankle kinetics,
contralateral stability, and leg length on stride length.
Significance was set at p < .05.
Main Results: None of the hip or knee kinematic measures
significantly predicted stride length. Only when combined
with leg length, ankle push-off energy and first double
support did the influence of hip and knee motion become
significant. The following equation was derived from the
patient data (p < .0454 for each variable; R2 = .739):
Stride length (cm) = -0.130 + 1.165(leg length) + 0.637(hip
arc) - 2.082(first double support) + 1.267(contralateral
ankle push-off energy) + 0.647(ankle push-off energy)
Only ankle push-off energy and leg length were predictive
of stride length in the normal subjects (p < .007 for each
variable; R2 = .832).
Conclusions: Stride length appears significantly related
to leg length, ankle push-off energy, first double support,
and hip arc. No other kinematic variables significantly
influenced stride length. Increased knee extension in
stance and terminal swing, often achieved with hamstring
lengthenings, did not appear to influence stride length.
Therefore surgical intervention to correct crouch may have
a limited effect on stride length.
Acknowledgment: We would like to thank Linda Smith for
assistance in data collection.
Gait and Posture, 7 (2) p.148, 1998
PREDICTORS OF STRIDE LENGTH BAREFOOT AND WITH ANKLE FOOT
ORTHOSES IN CHILDREN WITH CEREBRAL PALSY
Michael S. Orendurff, MS, James S. Chung, BS, Robin
Dorociak, BS, and Rosemary Pierce, PT Gait Analysis
Laboratory, Shriners Hospital for Children, Portland,
Oregon
Introduction
Previous studies have developed
statistical models
which use kinematic and kinetic variables to predict stride
length in children with cerebral palsy (Orendurff, et al.,
1997). These models suggest that leg length, contralateral
ankle push-off energy and first double support time account
for a majority of the variance in stride length in
barefoot, appliance-free gait. Increased knee flexion did
not appear to influence stride length to a great extent,
accounting for only 6% of the variance. It is our
observation that children often improve their stride length
when wearing ankle foot orthoses (AFOs), which often
restrict ankle motion and therefore push-off energy, but
may also improve stability in single limb stance. The
purpose of this study was to evaluate the effect of AFOs on
hip and knee motion, ankle kinetics and stride length in
free walking children with cerebral palsy.
Methods
Three cohorts were used in
this study. This first
was made up of 106 free walking children diagnosed with
cerebral palsy (CP) who received gait analysis as part of
their ongoing care at a regional children's hospital. We
retrospectively evaluated barefoot gait kinematics and
kinetics from a 6-camera Vicon 370 system with 2 AMTI force
plates. A second group, 37 similar patients with AFOs were
similarly evaluated. The third group containing 24
children with CP with matched free walking barefoot and AFO
gait data were used to validate these models. Variables
included leg length, (LL), stride length (SL), sagittal hip
arc (HA), peak knee extension in single limb stance (PKE),
peak knee flexion (PKF), peak knee extension in terminal
swing (PKEtsw), ankle push-off energy (A2E), contralateral
ankle push-off energy (A2Ecl), and first double support
time (DS1). These variables were extracted from the data
and analyzed using StatView to determine if different
strategies were utilized when the subjects were in braces
compared to barefoot gait (multiple linear regression) and
to determine which variables showed significant differences
between barefoot and brace gait (repeated measures ANOVA).
Results
For the barefoot free
walking subjects the
following model was derived: SL = 1.054 - 1.810(DS1) +
1.145(LL) + 1.052(A2Ecl) + 0.629(HA) + 0.468(A2E);
p < .0001, R2 = .749, n = 106.
For the braces free walking
group a slightly
different model emerged:
SL = 20.715 - 1.835(DS1) + 1.112(HA) - 0.857(PKEtsw) +
0.467(PKE) + 0.101(LL); p < .02, R2 = .831, n = 37.
When applying the barefoot
model to the barefoot
data from the matched group, ankle push-off energy (A2E)
was not significant (p = .6861) but all other variables
were significant (p < .04) and provided good fit for
predicting stride length (R2 = .822).
The brace data from the
matched group was fit to
the brace model derived above and peak knee extension in
single limb stance (PKE) was not significant (p =.0753),
but all other variables were significant (p <.006) and
yielded good fit for predicting stride length (R2 = .810).
The repeated measures ANOVA
showed that most
variables did not change significantly in brace versus
barefoot gait. These differences are summarized below.
Table 1. Mean ± SD for 8 variables in 24 subjects with
matched barefoot and braced gait data Variable
Barefoot
Braces
p value
Stride Length SL (cm)
93.4 ± 15.7 103.4 ± 18.0
<.0001*
Hip Arc HA (q°)
47.4 ± 8.1 48.3 ± 8.2
.5419
Peak Knee Ext PKE (q°)
3.7 ± 10.2 3.9 ± 4.0
.8632
Peak Knee Flex PKF (q°
53.2 ± 10.7 56.8 ± 9.6
.0271*
Peak Knee Ext Swing PKEtsw (q°)
19.7 ± 11.0 19.9 ± 13.0
.8979
Ankle Push-Off Energy A2E (J/kgbw)
8.0 ± 4.2 5.1 ± 2.6
.0003*
Contralateral Ankle Push-Off Energy A2Ecl (J/kgbw) 7.6 ± 3.6
6.6 ± 3.3 .2505
First Double Support Time DS1 (%)
14.9 ± 2.2 15.6 ± 3.1
.2187
Discussion
Clearly, braces improve
stride length significantly
in this group of 24 free walking children with cerebral
palsy. Of this cohort, all but 3 subjects showed increases
in stride length with bracing. There appeared no
consistent change in this group in the variables measured
that would account for the increase, that is no specific
improvement in a set of variables that would suggest one
strategy which is present for all members of the group. It
appears that braces improved stride length in each
individual by varied mechanisms.
It was theorized from the
model of stride length
that contralateral ankle push-off energy (A2Ecl) would
create acceleration, and stability in stance would provide
a means to apply this acceleration to the center of mass,
moving it forward. As expected ankle push-off energy was
significantly reduced with bracing but the theorized
concomitant improvements in stability--as quantified but
first double support time (DS1)--were not apparent in these
data. Due to different levels of involvement in this group
it is reasonable to expect varying responses to bracing.
Since all types of bracing were included in this cohort, a
more refined study to document attributes of several types
of braces has been undertaken. The data show that the
regression models for predicting stride length in barefoot
and braced walking have good fit, and that despite large
individual differences the mechanisms for achieving stride
length may be quite different in these two conditions. One
possibility is that in patients who are able to compensate,
decreases in stride length due to decrements in knee motion
may overcome by increased push-off in barefoot gait. With
bracing, when ankle motion and therefore push-off is
reduced, restrictions at the knee appear to have a greater
impact, especially PKEtsw, even though they account for
only about 14% of the variance in the braced trials. Based
on the highly variable nature of this population,
evaluation of each individual for specific causes of
abnormal gait is warranted before recommendations regarding
bracing are made.
Reference
Orendurff, MS, et al. Dev Med and Child Neuro, 39(S75):3,1997.
(Proceedings of the American Society of Biomechiancs
Conference, 1997)
STRIDE LENGTH CHANGES FOLLOWING SURGICAL HAMSTRING
LENGTHENINGS IN INDIVIDUALS WITH CEREBRAL PALSY
M. Orendurff, R. Pierce, R. Dorociak, M. Aiona
Gait Laboratory, Shriners Hospital for Children, Portland
Unit, Portland, Oregon USA 97210
INTRODUCTION
Individuals with cerebral palsy (CP) may develop muscular
contractures which can limit joint motion. For example,
tight hamstrings can lead to flexed-knee gait, making
ambulation difficult. If a contracture of the hamstrings
is sufficient to warrant intervention, a surgical
lengthening may be performed with the goal of increasing
knee extension and correcting the patient's crouch gait.
Since crouch gait is claimed to be associated with
decreased stride length (Root, 1992), a collateral goal of
this surgery is often to improve stride length.
REVIEW AND THEORY
Previous work in our laboratory (Orendurff, et al., 1997)
led to the development of a regression equation to predict
stride length in individuals with CP. An equation was
developed with 54 consecutive individuals (54 sides chosen
randomly) with CP (mean age 11.8 ± 4.2 years, range 20.5 to
4.7 years) who under went computerized gait study as part
of their ongoing care. From this group the following
equation was derived:
Stride length (cm) = -0.130 + 1.165(leg length) + 0.637(hip
arc) - 2.082(first double support) + 1.267(contralateral
ankle push-off energy) + 0.647(ankle push-off energy)
p <.0454, R2 = .739
This equation suggests that stride length is not associated
with restricted knee motion, or that other strategies may
be used to compensate for restricted knee motion. The goal
of the present study was to determine which variables
improved following hamstring lengthenings, and how these
changes were associated with increased stride length.
PROCEDURES
In this experiment, we retrospectively evaluated 15
consecutive individuals with crouch gait (15 sides chosen
randomly) to examine knee kinematics prior
to
and one year post hamstring lengthenings. These patients
received computerized gait analysis as part of their
ongoing care at a regional children's hospital. In
addition to hamstring lengthenings, 4 individuals also had
tendo achilles lengthenings, 3 had rectus femoris
transfers, 6 had adductor lengthenings, and 2 had iliopsoas
lengthenings. Barefoot, appliance-free gait data were
collected 4.2 ± 1.4 months prior to surgery and 14.0 ±
4.3
months following surgery, for a total of 18.1 ± 5.0 months
between gait studies. A six camera VICON 370 system with
two AMTI force plates was used to collect gait data, which
was processed using Vicon Clinical Manager (Oxford Metrics,
Oxford, England). Variables of interest were extracted and
analyzed using StatView (Abacus Concepts). Analysis of
variance was used to determine which variables showed
significant differences postop. In addition, multiple
linear regression was used to predict stride length based
on leg length (LL), sagittal hip arc (HA), first double
support period (DS1), contralateral ankle push-off energy
(A2Ecl), and ipsilateral ankle push-off energy (A2E) both
preop and postop. The p < .05 level of significance was
chosen.
RESULTS
Following hamstring lengthenings significant improvements
were seen in stride length (SL) total knee arc (KA), peak
knee extension (PKE) and peak knee extension in terminal
swing (PKEtsw) (p < .01). Peak knee flexion (PKF) showed
a
significant decrease following intervention (p < .01).
None of the variables from the multiple regression model to
predict stride length showed significant differences
following surgery, with the exception of leg length (LL)
which increased significantly (p = .0002) presumably due to
growth.
When applying the model to the preop patient data only
three variables proved significant (LL, HA and A2Ecl; p <
.02, R2 = .797). Postop, HA and A2Ecl dropped out of the
model and only LL, DS1 and A2E proved significant
predictors of stride length (p < .007, R2 = .917).
Preop
Postop p value
SL 95.8 ± 17.9
103.8 ± 18.0 .0056
KA 31.6 ± 15.4
40.5 ± 12.7 .0099
PKE 24.6 ± 9.7
8.1 ± 9.4 <.0001
PKEtsw 32.5 ± 7.7 17.9 ±
6.7 <.0001
PKF 56.2 ± 12.1
48.6 ± 9.5 .0050
Table 1. Knee kinematics following hamstring lengthenings
(mean ± SD).
Preop
Postop p value
LL 75.2 ± 12.1
78.7 ± 10.4 .0002
HA 41.8 ± 8.3
43.1 ± 6.4 .5368
DS1 13.7 ± 4.7
15.5 ± 2.4 .1300
A2E 6.4 ± 2.5
6.9 ± 2.1 .4442
A2Ecl 8.3 ± 4.4
7.6 ± 2.9 .5100
Table 2. Stride length prediction variables following
hamstring lengthenings (mean ± SD).
DISCUSSION
This group of individuals with CP who underwent hamstring
lengthenings to correct crouch gait showed significant
improvements in knee kinematics, including improvements in
knee extension in stance and in terminal swing. Stride
length significantly improved 8 cm postop for the group,
but individual results were highly variable. Six
individuals showed stride length increases of between 10
and 28 cm (Improved), five showed increases between 0 and 7
cm (Unchanged) and four showed decreases in stride length
of between 0 and 3 cm (Decreased) even though improvements
in knee kinematics were similar. Much of the overall
change in stride length for the group can be attributed to
changes in leg length due to growth between the preop and
postop evaluations: leg length alone can account for 40% of
the stride length variance.
Similar improvements in stride length were seen in a group
of 8 individuals with CP who had gait abnormalities, but
did not require intervention. They increased their
stride length 8 cm as well (p = .0064)
in 15.8 + 2.8
months, presumably due to growth.
Figure 1. Improvements in stride length following surgical
hamstring lengthenings.
In an effort to understand why some individuals improved
their stride length following intervention and others did
not, patient data was fitted to a model developed earlier.
Although this method does reveal significant results,
indicating that hip arc and contralateral ankle push-off
energy were important in determining stride length preop
and that the first double support period and ipsilateral
ankle push-off energy were important postop, interpretation
is complex. It seems unlikely that mechanisms to increase
stride length would be so different from preop to postop
without significant changes in the variables following
intervention.
By evaluating the three groups (Improved, Unchanged and
Decreased) using a Kruskal-Wallis test to determine which
variables were associated with improved stride length it
appears that improved sagittal hip motion was associated
with increased stride length postop (p< .03).
This suggests that individuals who were able to compensate
for their knee flexion preop by increasing hip motion did
not show large improvements in stride length, whereas
individuals who had knee flexion and were unable to
compensate with sagittal hip motion showed large changes in
stride length following intervention. Further study of
this highly variable group is needed before the effects of
improved knee extension on stride length are known.
REFERENCES
Orendurff, M.S., et al., Dev Med Child Neurol, in press,
1997.
Root, L., in M. Sussman, Ed., The Diplegic Child, 309-315,
AAOS, Rosemont, IL., 1995.
I would be interested to
hear other comments about
this child, and stride length in general.
Sincerely,
Michael
Michael Orendurff, MS
Clinical Biomechanist
Motion Analysis Laboratory
Portland Shriners Hospital
First, I want to make sure we're all talking about the same thing. Michael's abstracts (thanks very much for them - interesting reading!) repeatedly refer to "contralateral stride length", which, to me, is an oxymoron. Stride length refers to the whole gait cycle, so it is neither refers to left/right, or ipsilateral/contralateral. I presume you are meaning the stride-length that happened to be measured at the time? Perhaps you could clarify this, Michael?
That said, looking at the figures, they do seem to confirm a strong correlation between stride length and push-off energy in normal children. The correlation with leg length is, of course, also expected, and really we ought to be normalizing for this. As suggested by At Hof - see /faq/normalisation.html - we ought to divide stride length (SL) by leg-length (l), and energy (E) by m.g.l where m is body mass and g gravity. That said, nobody ever seems to do it, so I guess SL/l and E/(m.l) will suffice.
The correlation with double support again makes sense, since the longer the foot is on the ground the longer the time for push-off and the greater the work done.
One more thing I would like to clarify. In the discussion of the Gait and Posture (7 (2) p.148, 1998) paper, you say contralateral ankle push-off energy (A2Ecl) would create acceleration, and stability in stance would provide a means to apply this acceleration to the center of mass, moving it forward. I disagree that this is a summary of current-thinking. As I understand it (At Hof again!, J Biomech 16: 523 - 537, 1983, I think) the energy from push-off is used to propel the swing leg forward - very little (if any) of this energy flows through the hip joint to the trunk (whose energy is almost constant). In other words. push-off is responsible for leg propulsion, not trunk progression. Again, I think its time we agreed on this - I have many people wrongly assert that push-off pushes the trunk forwards.
As far as the situation in CP is concerned, all of this may be irrelevant. We are dealing with a different system, and your results seem to show this. I would contend, though, that this particular patient (with a minor involvement) would behave similarly to the normal child. I will ask the PT in charge to try an AFO to see if stride length improves.
I'm glad you asked about the endpoint of treatment. I was actually going to ask about this - I confess that I don't know what is normally expected from Botox. I think sometimes, it is hoped that the temporary relaxation during the 3 month period when it is effective will have some carry-over effect. I guess the alternatives for the long-term are serial-casting or surgery. I do think Botox is an ideal method for assessing the likelihood of success of these more time-consuming/invasive procedures. Perhaps someone with more experience can tell us what usually eventuates after a successful course of Botox?
Chris
--
Dr. Chris Kirtley
Dept. of Rehabilitation Sciences
The Hong Kong Polytechnic University
--M
As often happens, the discussion seems to have focussed in on a more
basic biomechanical point: this time the puted relationship between
stride length and push-off. I'm happy to have this debate as I think
it is time we sorted out this controversial topic once and for all!
First, I want to make sure we're all talking about the same thing.
Michael's abstracts (thanks very much for them - interesting
reading!) repeatedly refer to "contralateral stride length", which,
to me, is
an oxymoron. Stride length refers to the whole gait cycle, so it
is
neither refers to left/right, or ipsilateral/contralateral. I presume
you are
meaning the stride-length that happened to be measured at the time?
Perhaps you could clarify this, Michael?
This is news to me. I never referred to it as far as I
know. Having searched through the abstracts I can only
think that a table got pasted in with incorrect tabs and
the "contralateral" somehow got placed next to stride
length. I never got my own posting so perhaps you could
forward it to me Chris and and I'll try to explain it!?
That said, looking at the figures, they do seem to confirm a strong
correlation between stride length and push-off energy in normal
children. The correlation with leg length is, of course, alsoexpected,
and really we ought to be normalizing for this. As suggested by
At Hof -
see /faq/normalisation.html - we
ought to divide stride length (SL) by leg-length (l), and energy
(E) by
m.g.l where m is body mass and g gravity. That said, nobody ever
seems to
do it, so I guess SL/l and E/(m.l) will suffice.
The correlation with double support again makes sense, since the
longer the foot is on the ground the longer the time for push-off
and the
greater the work done.
In CP this might mean the inability to ever get any
push-off from the ankle. The most involved have the longed
double support.
One more thing I would like to clarify. In the discussion of the
Gait
and Posture (7 (2) p.148, 1998) paper, you say "contralateral ankle
push-off energy (A2Ecl) would create acceleration, and stability
in
stance would provide a means to apply this acceleration to the center
of mass, moving it forward". I disagree that this is a summary of
current-thinking.
It was a summary of my current thinking, not any one elses.
As I understand it (At Hof again!, J Biomech 16: 523 - 537, 1983,
I think)
the energy from push-off is used to propel the
swing leg forward - very little (if any) of this energy flows through
the
hip joint to the trunk (whose energy is almost constant). In other
words.
push-off is responsible for leg propulsion, not trunk progression.
Again, I think its time we agreed on this - I have many people
wrongly assert that push-off pushes the trunk forwards.
(I'm unclear as to whether we're discussing "what it does
in normal" or "what is possible".) First the trunk's
velocity is not constant in gait so I doubt it energy is
constant. If the A2 power burst does not flow upward to
some extent then why are there coincident power bursts
during pre-swing at the knee and hip? What about in
jumping, where the ankle push-off energy is huge--doesn't
that accelerate the trunk? If the ankle power only
accelerates the leg into swing then why does swing (and
stride length) increase with a rigid AFO? My assertion is
that the position of the ankle relative to the position of
the COM will accelerate the COM in a forward direction.
Some if the ankle power may also contribute to swing the
leg.
As far as the situation in CP is concerned, all of this may be
irrelevant. We are dealing with a different system, and your results
seem to show this. I would contend, though, that this particular
patient (with a minor involvement) would behave similarly to the
normal
child. I will ask the PT in charge to try an AFO to see if stride
length
improves.
I'm wondering is her contralateral vaulting will improve
with a AFO? She is very young and has a fair amount of
variability so perhaps only time will tell.
I'm glad you asked about the endpoint of treatment. I was actually
going to ask about this - I confess that I don't know what is normally
expected from Botox.
I wish I could say the results were generally dramatic.
Perhaps in this case they will be, since you can give a
relatively large does to one muscle rather than spreading
it around. In the US our dosage guidelines appear to be
lower than Europe, and we rarely get much effect if we
spread the dosage out to 4 muscles (say hams and gastrocs
bilaterally). Do you use a local? a general? We just use a
hypnotic and the kids don't like the procedure one bit, but
never remember it. Perhaps the parents need some too. I'm
not sure I could comfort my child through the procedure
every three months for ???.
I think sometimes, it is hoped that the temporary
relaxation during the 3 month period when it is effective will have
some carry-over effect. I guess the alternatives for the long-term
are
serial-casting or surgery. I do think Botox is an ideal method for
assessing the likelihood of success of these more
time-consuming/invasive procedures. Perhaps someone with more
experience can tell us what usually eventuates after a successful
course of
Botox?
With my experience either continuous injections or surgery.
Others at this institution are in the middle of a
multi-center clinical trial looking at botox and serial
casting so perhaps I'll be pleasantly surprised by the
results.
Michael Orendurff, MS
Clinical Biomechanist
Motion Analysis Laboratory
Portland Shriners Hospital
Lastly I find it a very useful tool in the very young child when
spasticity is interfering with weight bearing function or because
of
excessive hip adduction. Reducing the spasticity in this kids
may
reduce the degree of deformities long term and initiate walking at
an earlier
stage in their development.
MossRehab Gait & Motion Analysis Laboratory
chris morris
orthotist
oxford
uk