Dislocation of the hips
with bilateral spastic cerebral palsy (CP). This is due to
in children with
bilateral spastic
cerebral palsy,
1985–2000
R E Morton;
B Scott;
V McClelland;
A Henry, Ronnie MacKeith Child Development Centre and
Derbyshire Children’s Hospital, Derby, UK.
*Correspondence to first author at Ronnie MacKeith Child
Development Centre and Derbyshire Children’s Hospital,
Derby, UK.
E-mail: richard.morton@derbyhospitals.nhs.uk
The aim of this study was to assess the rate of hip dislocation
at different ages in children with bilateral spastic cerebral
palsy attending special schools in southern Derbyshire, UK,
between 1985 and 2000. The medical notes of 110 individuals
(68 males, 42 females) were obtained. They were divided into
four groups according to the Gross Motor Function
Classification System (GMFCS). We determined whether or
not their hips were dislocated at the ages of 5, 10, and 15
years, and the kind of surgery performed in each case. The
percentage of individuals with one or both hips dislocated
increased with age and with severity of disease. Of those in
GMFCS Level II (n=18), none had dislocations; Level III
(n=16), none had dislocations at ages 5 and 10, but 11% had
by the age of 15; Level IV (n=35), 8% had dislocations by age
5, 19% by age 10, and 30% by age 15; Level V (n=41), 22%
had dislocations by age 5, 48% by age 10, and 50% by age 15.
Forty-two per cent of individuals with hip dislocation had not
had previous preventive surgery. Twenty-one per cent of hips
operated on still proceeded to dislocation. We conclude that
there was a high rate of hip dislocation, especially in GMFCS
groups Levels IV and V, and that this often occurred very
early. Preventive surgery avoided dislocation in many
children. However, orthopaedic referral was often not made
before dislocation was discovered, or the referral was made
too late for surgery on soft tissue to be successful. These
results may be compared with those from current
programmes of hip management, involving radiological
surveillance and early use of conservative and surgical
interventions.
Downloaded from https://www.cambridge.org/core. University of Calgary Library, on 06 Nov 2017 at 13:14:12, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0012162206001174
The hip joint is at risk of abnormal development in children
unusual forces being applied to the joint from spastic muscles,
causing hip adduction and flexion, combined with inade-
quate weight bearing. This may lead to hip dislocation, caus-
ing pain, problems in seating, and perineal hygiene. It also
contributes to the development of scoliosis (Kalen and Bleck
1985). Current practice is for close surveillance of the hips of
those at risk (Scrutton et al. 2001) and early intervention
with postural management and surgery (Reimers 1980,
Scrutton 1989). To evaluate the outcome of this manage-
ment in the future, it is important to fully document results
from previous approaches.
In a group of severely affected children with CP, 64% had
hip dislocation (Samilson et al. 1972). In contrast, much
lower rates of only 5% (Fujiwara et al. 1976) and 6% (Beals
1966) have been reported in children who were more active.
These studies were conducted in special centres with select-
ed patients who may not have accurately reflected the popula-
tion of children with bilateral CP. It is also unclear from some
of the studies how long follow-up was maintained, which is
important as hip dislocation can occur into the late teens or
early adulthood. In one of the few epidemiological studies of
hip dislocation that traced the majority of severely affected
individuals with bilateral CP throughout childhood, 47 out of
327 (14%) had dislocated hips (Cooke et al. 1989). However,
this is difficult to compare with modern outcome studies
because a system of classifying children with CP accurately
according to severity was not available at the time of the study.
We studied a birth cohort of children with bilateral CP in
the area of the old Southern Derbyshire Health Authority, UK,
who were treated in the conventional manner before mod-
ern methods of hip management were available. They were
managed by a multidisciplinary team, with orthopaedic surgery
taking place at the Derbyshire Children’s Hospital mainly by
one orthopaedic surgeon qualified to work with children
and adults. Children were classified retrospectively accord-
ing to severity of disease by using the Gross Motor Function
Classification System (GMFCS; Palisano et al. 1997).
Method
We obtained the names of pupils with bilateral spastic CP who
had attended four special schools in the area of the old
Southern Derbyshire Health Authority between 1985 and
2000. This amounted to most children with bilateral CP born
between 1971 and 1995, except those who did not attend one
of the special schools in the period and a few who may have
died before 1985. There was incomplete follow-up to the age
of 15 years of those still under this age when recording was
stopped in 2000. The hospital notes were acquired and the
diagnosis of bilateral spastic CP confirmed, excluding those
wrongly identified. The status of the hips was determined as
whether dislocated or not, at 5, 10, and 15 years, with details
of hip surgery. Hip dislocation was defined as a hip outside
the acetabulum and confirmed to have a migration over 100%
on X-ray. Ethical permission was obtained from the local
Southern Derbyshire Ethics Committee.
PARTICIPANTS
From 1971 to 1995, when the participants were born, the aver-
age birth rate for the hospital was 5500 per year and the mean
rate of CP was 0.22%. Of these, 51% could be expected to
Developmental Medicine & Child Neurology 2006, 48: 555–558 555
 develop bilateral spastic CP (Hagberg 1993). Therefore, the
total number of participants expected from this cohort was
approximately 154. Four were known by the authors to have
died in the first year of life before they could first be recorded at
age 5 years in 1985, leaving around 150 to be traced. We deter-
mined the names of 127 with bilateral spastic CP, but for 17 the
notes were unobtainable. This left the notes of 110 patients for
analysis (68 males, 42 females), i.e. 73% of the 150 expected.
MANAGEMENT
The care provided for these patients was typical of the peri-
od. Supportive seating was provided for most of the waking
hours, and a handful were also managed in a hip brace for
part of the day. The hip surgery performed was preventive,
reconstructive, and salvage (Miller et al. 1995). Preventive
surgery was to the soft tissues only, consisting of open adduc-
tor tenotomy, psoas tenotomy, and obturator neurectomy,
apart from four children from GMFCS Level V who could
only tolerate a short general anaesthetic and who had percu-
taneous adductor tenotomies alone. This was performed
when the hips were clinically felt to be severely adducted,
confirmed by an X-ray showing migration approaching 50%.
Reconstructive surgery was performed for five dislocated
hips; this was successful in two. Two patients with bilateral
hip dislocation had the heads of both femora removed, which
was successful in eventually reducing their pain and improv-
ing seating posture.
Results
HIP DISLOCATION
Results were arranged according to GMFCS groups and are
given in Table I and Figure 1. Hip dislocation occurred in 30
out of 110 (27%) children; in eight children (7%) this was
bilateral. Therefore, from a total 220 hips, 38 were dislocated.
PREVENTIVE SURGERY
Hip dislocation related to surgery for different GMFCS groups
is shown in Figure 2.
GMFCS Level II
No hips were dislocated. Six hips were operated on.
GMFCS Level III
One hip dislocated. Thirteen hips were operated on, one later
becoming dislocated after two preventive operations at age 4
and 6 years. Mean age of surgery on hips without later dislo-
cation was 7 years 1 month.
GMFCS Level IV
A total of 11 hips dislocated, eight without previous operation.
Twenty hips were operated on, with three later becoming
dislocated (15%). Mean age of surgery on hips with later hip
dislocation was 6 years 1 month, and without dislocation
6 years 7 months.
GMFCS Level V
A total of 26 hips dislocated, 16 without previous operation.
A total of 30 hips were operated on, with 10 later dislocating
(33%). Mean age of surgery with later hip dislocation was 4
years 2 months, and without dislocation 5 years 6 months.
Overall, the mean age of unsuccessful surgery for hips
which later became dislocated was 4 years 11 months, and
successful surgery for those remaining in joint was 6 years
(p=0.15, not significant).
Discussion
This study was performed before the local database for chil-
dren with CP was established. We were only able to examine
the notes of 73% of the estimated number from the birth
cohort studied. It is likely that most of those from GMFCS
groups Levels IV and V would have been included as they
would have attended one of the special schools. However, sev-
eral of those with less severe CP may have attended main-
stream school and been missed, weighting the total sample
towards those with more severe disease. This occurred
increasingly over the period covered by the 25 year birth
cohort and explains the high rate of hip dislocation in this
series of 27% overall, compared with 14% in a comparable epi-
demiological study (Cooke et al. 1989). Despite this skewed
overall representation of patients with CP, valid conclusions
may be drawn about the prevalence of hip dislocation at differ-
ent ages for each separate group according to the GMFCS.
These results show an alarmingly high rate of hip dislocation,

Table I: Hip dislocation in children in different GMFCS groups by 5, 10, and 15 years of age

GMFCS Level II GMFCS Level III

Age, y n n (%) with n (%) with bilateral n n (%) with n (%) with bilateral
dislocation dislocation dislocation dislocation

5 18 0 0 16 0 0
10 18 0 0 16 0 0
15 18 0 0 9 1 (11) 0
GMFCS Level IV GMFCS Level V
n n (%) with n (%) with bilateral n n (%) with n (%) with bilateral
dislocation dislocation dislocation dislocation

5 35 3 (8) 1 (3) 41 9 (22) 4 (10)

10 35 7 (19) 1 (3) 33 16 (48) 4 (12)
15 27 8 (30) 2 (7) 24 12 (50) 3 (13)

Not every child was followed up for 15 years so the three cohorts for each group are different. GMFCS IV, four children dislocated between 5 and 10
years, all unilaterally; and two between 10 and 15 years, one bilaterally. GMFCS V, nine children dislocated between 5 and 10 years, two bilaterally
and two between 10 and 15 years, both unilaterally. GMFCS, Gross Motor Function Classification System; n, number followed up.

556 Developmental Medicine & Child Neurology 2006, 48: 555–558

Downloaded from https://www.cambridge.org/core. University of Calgary Library, on 06 Nov 2017 at 13:14:12, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0012162206001174
 especially in non-mobile children in GMFCS groups Levels
IV and V, giving a combined figure of 39% with a dislocated
hip by age 15 years. We were also surprised by the number
of these children with dislocations at only 5 years of age,
amounting to 12 out of 30 children with dislocations (40%).
This figure is slightly elevated because some children were
not followed up until 15 years, but is comparable to a figure
of 32% in the study by Cooke et al. (1989). Other studies in
similar children have suggested that hips tend to dislocate
later, towards the age of 10 years, with an average rate of
migration of only 7 to 9% per annum (Vidal et al. 1985). It
appears, therefore, that epidemiological studies like ours
may give a fuller picture of the actual rate of dislocation in
some children, including those who dislocate early and who
may not be referred to tertiary orthopaedic centres.
We found that 63% of those with hip dislocation from groups
IV and V did not have the chance of preventive surgery before-
hand. Of those who did have preventative surgery from these
groups, 21% subsequently dislocated. Surgical procedures
used were typical of the period, and some now question
whether obturator neurectomy was necessary.
The timing of surgery is very important. We found that the
mean age for surgery in hips that later became dislocated
tended to be less than for those that remained stable. This
was presumably because they were discovered at an earlier
stage, but still not soon enough for successful results. In a
series reported by Reimers (1980), in which surgery was gen-
erally performed much earlier, the best results were pro-
duced before the age of 4 years. He therefore recommended
early surgery when the hip migration percentage reached
about 33%; he pointed out that if left over the age of 5 years,
soft tissue surgery alone was insufficient and bony proce-
dures should be considered.
We found that surgical procedures to reduce the hip after
it has dislocated are often unsuccessful, even when only the
most able and potentially mobile patients are selected. The
procedures involved are complex and others have found that
60
Children with hip dislocation (%)
50
40
30
20
10
0 0
Level II Level III Level IV
GMFCS group
Figure 1: Hip dislocation in children with cerebral palsy
according to Gross Motor Function Classification System
(GMFCS).
Downloaded from https://www.cambridge.org/core. University of Calgary Library, on 06 Nov 2017 at 13:14:12, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0012162206001174
as few as a third of hips subsequently stay in joint (Samilson
et al. 1972). An alternative is to remove the head of the femur
to allow adequate hip abduction. However, this is a very painful
procedure, which was performed in only two of our patients.
Recently we have found by parental report that injections of
botulinum toxin to the adductors relieves the pain in dislo-
cated hips for 5 months or more (Campbell et al. 2005),
which is particularly helpful for those who have become too
unwell for surgery under a general anaesthetic.
In the period covered by this study, significant hip sublux-
ation tended to be discovered clinically, usually late, and
then confirmed by X-rays. Scrutton and Baird (1997) showed
that over 80% of hips with a migration percentage over 33%
at 30 months needed treatment by 5 years. They recom-
mended that all children with bilateral CP unable to walk
more than 10 steps alone at 30 months should have a hip X-
ray in the recommended position to measure the migration
percentage, or before if clinically indicated. Further X-rays
were performed according to the results, initially every 6
months for those with the most severe disease. They further
suggested referral to an orthopaedic surgeon when the
migration percentage exceeded about 30%, emphasizing the
need to centralize the hip by the age of 4 years while there is
still time for the acetabulum to develop.
There is some debate about the relative importance of
preventive surgery and conservative measures. Many would
underline the importance of 24-hour postural management
(Pountney et al. 2001) involving correct seating, hip abduc-
tion braces in the day, and abduction beds at night, with use
of a standing frame and passive stretching. Botulinum injec-
tions to the adductors are also helpful (Heinen X, personal
communication, 1995). Abduction beds have been shown to
reduce subluxation (Hankinson and Morton 2001). It seems
sensible to start these early to encourage development of the
acetabulum, and in those with severe spasticity at particular
risk of subluxation, even before the first X-rays at 18 to 24
months. We would now start postural therapy anyway if the
70
60
Hips dislocated (%)
50
Hips
5y 40 dislocated
10y after surgery
30 Hips
15y dislocated
without
20 surgery
10
Level V Level II Level III Level IV Level V
GMFCS group
Figure 2: Hip dislocation related to surgery for different
Gross Motor Function Classification System (GMFCS)
groups.
Hip Dislocation in Children with Cerebral Palsy R E Morton et al. 557
 migration percentage exceeded 15 to 20% in the first 5 years,
referring for orthopaedic opinion in the usual way over 30%.
Postural management alone is unlikely to be sufficient for
many children, but as Scrutton (1989) says, this is not in com-
petition with surgery, it should be complementary to it.
Conclusions
Our results show that: (1) hip dislocation was common in
children with bilateral CP, especially in those most severely
affected in GMFCS groups Levels IV and V, some of whom dis-
located before the age of 5 years; and (2) preventive soft tis-
sue surgery, consisting of psoas and adductor tenotomy with
obturator neurectomy, reduced the incidence of hip disloca-
tion when it was performed; however, this was often done
too late or not at all. It is hoped that current management of
at-risk hips will be more successful, consisting of early identi-
fication of subluxation by X-ray surveillance, and early inter-
vention with conservative measures and surgery together.
Our findings can be compared with future, similar, popula-
tion-based studies to see if this proves to be the case.
DOI: 10.1017/S0012162206001174
Accepted for publication 13th May 2005.
Acknowledgements
We thank the nursing sisters in the special schools, including
Judy Paillin, Hazel Massey, and Christine Tipping.
References
Beals RK. (1966) Spastic paraplegia and diplegia; an evaluation of
non-surgical and surgical factors influencing the prognosis for
ambulation. J Bone Joint Surg Am 48: 827–846.
Campbell V, Broderick M, Morton RE. (2005). Treating painful hips
International Child Neurology Association (ICNA)
ICNA is a respected international group of child neurologists who wish to
advance the care of children with neurological problems throughout the
world.We invite all neurologists to join the association. Membership of a
regional child neurology association is a guarantee of becoming an ICNA
member. Membership benefits include:
Being part of an international network of child neurologists
Newsletter twice a year
The International Review of Child Neurology (IRCN) book series with
one free volume each year
Directory of members
Reduced fees for ICNA meetings
Participation in ICNA projects
For further information about categories of membership, application
procedures, dues, and the list of IRCN titles, go to:
http://www.child-neuro.net/icna.htm
558 Developmental Medicine & Child Neurology 2006, 48: 555–558
Downloaded from https://www.cambridge.org/core. University of Calgary Library, on 06 Nov 2017 at 13:14:12, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0012162206001174
in children too vulnerable to undergo surgical procedures.
Dev Med Child Neurol 98 (Suppl.): 21–22. (Abstract)
Cooke PH, Cole WG, Carey RPL. (1989) Dislocation of the hip in
cerebral palsy: natural history and predictability. J Bone Joint
Surg Br 71: 441–446.
Fujiwara M, Basmajian JV, Iwamoto M. (1976) Hip abnormalities
in cerebral palsy: radiological study. Arch Phys Med Rehabil
57: 278–281.
Hagberg G. (1993) The origins of cerebral palsy. Recent Adv Pediatr
11: 67–83.
Hankinson J, Morton RE. (2001) Use of a lying hip abduction system
in children with bilateral cerebral palsy: a pilot study. Dev Med
Child Neurol 44: 177–180.
Kalen V, Bleck EE. (1985) Prevention of spastic paralytic dislocation
of the hip. Dev Med Child Neurol 27: 17–24.
Miller F, Dabney KW, Rang M. (1995) Complications in cerebral palsy
treatment. In: Epps CH Jr, editor. Complications in Pediatric
Orthopedic Surgery. Philadelphia, PA: JB Lippincott.
Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B.
(1997) Development and reliability of a system to classify gross
motor function in children with cerebral palsy. Dev Med Child
Neurol 39: 214–223.
Pountney T, Green E, Gard P, Mandy A, Nelham R. (2001)
Retrospective analysis of hip migration in cerebral palsy, 2001.
Dev Med Child Neurol 42 (Suppl.): 85. (Abstract)
Reimers J. (1980) The stability of the hip in children: a radiological
study of the results of muscle surgery in cerebral palsy. Acta
Orthop Scand Suppl 184: 1–100.
Samilson RL, Tsou P, Aamoth G, Green WM. (1972) Dislocation
and subluxation of the hip in cerebral palsy. J Bone Joint Surg
54: 863–873.
Scrutton D. (1989) The early management of hips in cerebral palsy.
Dev Med Child Neurol 31: 108–116.
Scrutton D, Baird G. (1997) Surveillance measures of the hips of
children with bilateral cerebral palsy. Arch Dis Child 56: 381–384.
Scrutton D, Baird G, Smeeton N. (2001) Hip dysplasia in bilateral
cerebral palsy; incidence and natural history in children aged 18
months to 5 years. Dev Med Child Neurol 43: 586–600.
Vidal J, Deguillaume P, Vidal M. (1985) The anatomy of the
dysplastic hip in cerebral palsy related to prognosis and
treatment. Int Orthop 9: 105–110.