Predictive factors in open myelomeningocele with special reference to sensory level.
ABSTRACT A total of 113 cases of open myelomeningocele operated on shortly after birth were followed up and the 80 survivors (71%) were assessed one and a quarter to seven and a half years later. Their disability was classified in terms of mobility, intelligence, continence, and major complications; these when combined provided an assessment of overall disability. The overall disability of the survivors was minimal in 6%, moderate in 40%, severe in 39%, and very severe in 15%.A number of clinical features present at birth were analysed for their predictive value. Of these the sensory level, which frequently differed from both external and radiological levels of the lesion, correlated with the outcome in terms of mobility, intelligence, continence, major complications, and overall disability; and also with deaths caused by renal failure.A policy of confining operation to those patients with a reasonable chance of achieving independence would involve selecting for treatment a minority of all infants born with open myelomeningocele.
Article: Fetal surgery for myelomeningocele.[Show abstract] [Hide abstract]
ABSTRACT: Myelomeningocele is a devastating birth defect affecting a significant number of live births worldwide. Prenatal repair of myelomeningocele has been performed in the United States for 5 years with mixed results. The initial intent was to preserve distal neurological function by covering the exposed spinal cord. Although there has been relatively little effect on distal sensorimotor function, prenatal repair serendipitously led to an apparent reduction in hindbrain herniation and a possible decreased need for ventriculoperitoneal shunting. The long-term clinical consequences of these findings are not clear. What is clear, however, is that further study in the form of a prospective, randomized trial is mandatory.Current Opinion in Obstetrics and Gynecology 04/2001; 13(2):215-22. · 2.64 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: c1 Dr SN Sturgiss, Department of Fetal Medicine, Leazes Wing, Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP, UKFetal and Maternal Medicine Review 10/1995; 7(04):235 - 249.
- Advances in Pediatrics 08/2014; 61(1):33–74.
BRITISH MEDICAL JOURNAL
27 OCTOBER 1973
on its care of a child with a hopeless prognosis as on its suc-
cess with the survivors.
The feasibility and desirability of treatment of a grossly
handicapped infant, especially in the neonatal period, is one of
the most difficult decisions in paediatrics. Many factors are
involved and the varying attitudes of individual surgeons and
physicians and the pressures from parents accentuate the diffi-
culties. For centuries the mainstream of medical opinion has
abhorred the extremes and has trod the middle path. This
was expressed in classical form by the nineteenth century
English poet Arthur Clough (1819-61):
Thou shalt not kill; but need'st not strive
Officiously to keep alive.
This attitude has been re-enforced by a letter written by
Cardinal Jean Villot (1970), Vatican Secretary of State, to the
Federationof Catholic Medical Associations.
After stressing the sacred character of life that forbids a doc-
tor to kill and that imposes on him the duty to use all the
resources of his profession to fight against death, the Cardinal
writes: "But for all that, a medical man does not have to
There will always be differences of opinion between in-
dividual physicians and surgeons whether in a particular child
active treatment is justified or not, but we believe that com-
plete rejection and complete acceptance for treatment in spina
bifida are both untenable-the former denies active treatment
where the prognosis and the results of treatment are good
and gives the survivors less than an optimum future; the latter
all the techniques of survival offered by a constantly
keeps alive grossly handicapped children. Reasonably reliable
objective criteria are now available for selection for treatment
-their application has a marked effect on the quality of
We wish to record our thanks to the following members of the
spina bifida team who have worked with us in the clinic since 1961;
Mr. R. S. Hooper, neurosurgeon, and his staff; Mr. P. F. Williams
and Mr. M. B. Menelaus, orthomeedic surgeons; Dr. E. Wann,
psychiatrist; Miss Ruth Magnus, assistant urologist; and a num-
ber of medical social workers, appliance sisters, and paramedical
staff. All have contributed in various ways and have given freely
of their knowledge and experience. The teachers in the preschool
centres and snecial schools for the physically handicapped and,
not least, the parents of the children have also provided essential
information. To all we are greatly indebted.
Hagburg, B., and Siorgen, I. (1966). American3Journal of Diseases ofChildren,
Hide, D. W., Williams, H. P., and Ellis, H. L. (1972). Developmental
Medicine and Child Neurology, 14, 304.
Ingram, T. T. S., and Naughton, J. A. (1962). Developmental Medicine and
Child Neurology, 4, 287.
Laurence, K. M., and Coates, S. (1962). Archives of Disease in Childhood,
Lorber, J. (1971). Developmental Medicine and Child Neurology, 13, 279.
Lorber, J. (1972). Archives of Disease in Childhood, 47, 854.
Matson, D. D. (1969). Neurosurgery of Infancy and Childhood, 2nd edn.,
p. 36. Springfield, Ill., Thomas.
Sharrard, W. J. W., Zachary, R. B., Lorber, J., and Bruce, A. M. (1963).
Archives of Disease in Childhood, 38, 18.
Smith, E. D. (1965). Spina Bifida and the Total Care of Spinal Myelo-
meningocele, p. 75. Springfield, Ill., Thomas.
Smith, E. D. (1972). Journal of Pediatric Surgery, 7, 1.
Villot, Cardinal Jean (1970). L'Ossevatore Romano, Oct. 12-13.
Predictive Factors in Open Myelomeningocele with Special
Reference to Sensory Level
GILLIAN HUNT, WALPOLE LEWIN,JOHN GLEAVE, DOUGLAS GAIRDNER
British Medical Journal, 1973, 4, 197-201
A total of 113 cases of open myelomeningocele operated
on shortly after birth were followed up and the 80 sur-
vivors (71%) were assessed one and a quarter to seven
and a half years later. Their disability was classified in
terms of mobility, intelligence, continence, and major
assessment of overall disability. The overall disability
of the survivors was minimal in 6%, moderate in 40%,
severe in 39%, and very severe in 15%.
A number of clinical features present at birth were
analysed for their predictive value. Of these the sensory
level, which frequently differed from both external and
radiological levels of the lesion, correlated with the
outcome in terms of mobility, intelligence, continence,
major complications, and overall disability; and also
with deaths caused by renal failure.
Addenbrooke's Hospital, Cambridge CB2 2QQ
GILLIAN HUNT, M.B., D.C.H., Research Assistant
WALPOLE LEWIN, M.S., F.R.C.S., Neurological Surgeon
JOHN GLEAVE, F.R.C.S., Neurological Surgeon
DOUGLAS GAIRDNER, D.M., F.R.C.P., Paediatrician
A policy ofconfining operation to those patients with a
reasonable chance of achieving Independence would
involve selecting for treatment a minority of all infants
born with open myelomeningocele.
In the past few years there has been a change in the climate of
opinion, away from routine operative treatment in all cases of
myelomeningocele (Lorber 1971, 1972). At the same time it has
been confirmed that the most severe cases, if not operated on,
rarely survive for long-some 90% dying before the end of the
first year (Hide et al., 1972; Smith and Smith, 1973; Stark and
Drummond, 1973). A policy of selection of cases for operation,
therefore, seems to be indicated and attention needs to be
directed towards criteria by which cases likely to benefit by
surgery-in the sense of achieving a reasonable degree of
independence-can be more accurately identified at the time
of birth or soon after.
In this paper some predictive factors are examined. The
social consequences of this condition will be discussed in a
Subjects and Methods
This survey covers 113 consecutive cases of open myelomenin-
gocele which were operated on at Addenbrooke's Hospital
between 1 July 1963 and 1 January 1971 within 48 hours of
birth. It is an extension of the series described by Brocklehurst
et al. (1967). Simple meningoceles and late referrals are not
This survey was made between March 1971 and March 1972,
at which time there were 80 survivors aged 1 year 3 months to
7 years 8 months. Thirty-three children had died, 23 of these
(20% of all those operated on) before their first birthday. All
of the 80 survivors were located and visited, with the exception
of three who were too far away to visit and on whom medical
and school reports were obtained.
At the domiciliary visit a detailed history was obtained and a
intelligence an I.Q. was obtained on all the children over
5 years old through the co-operation of the school medical
service. For the children under 5 years old a developmental
assessment was made at the home visit. This was based on the
method used by Sheridan (1960) and Egan et al. (1969) but
modified to allow for the child's physical disabilities. No attempt
was made to obtain a precise score but merely to place the child
in one of four categories: above average, average, below average,
or severely retarded. The schools were also visited and the
records of the various hospitals attended were studied.
From the data on mobility, intelligence, and continence, com-
bined with any complication the severity of which materially
interfered with the child's daily life, a composite assessment of
overall disability was reached for each child. This overall
disability, as it affected the child's daily activities and related to
prospects of future independence, seemed to us to be the most
meaningful way to appraise a child with multiple handicaps.
Each of the 80 children was thereby placed in one of four
categories, defined as follows.
appliances and had a normal range of intelligence. All but one
Moderate Disability (32 cases).-All could walk but none
could run, some could walk only a few steps, nearly all had
calipers. One-third had low intelligence but none were severely
Severe Disability (31 cases).-None could walk 20 yards
(18 metres), some could not walk at all. Many had severe
kyphosis. There were all grades of intelligence including five
severely retarded. Only one child was continent.
Very Severe Disability (12 cases).-Ten of these could not
walk at all. All but one were severely retarded. Two were totally
blind, three were blind in one eye. All were incontinent.
The category of disability of the 80 survivors is shown in
OUTCOME RELATED TO PREDICTIVE FEATURES
Features noted at birth were now correlated with the overall
disability. Data were taken from the detailed notes recorded
during the original admission for closure, and generally made by
a senior member of the neurosurgical staff. The assessment at
follow-up was made by a single observer (G.H.). We have
analysed the outcome related to six predictive features: height of
sensory level; presence or absence of neural plaque, combined
with sensory level and separately; motor activity of the legs
in terms of quadriceps function; hydrocephalus
external level of the lesion; and results of infection.
The sensory level is taken as the first dermatome of normal
sensation above the area of anaesthesia. Sensory testing was a
BRITISH MEDICAL JOURNAL
27 OCTOBER 1973
routine part of the preoperative examination of every child.
The technique appears at first sight to be crude but in
experienced hands has proved to be remarkably accurate in the
long term. The resting baby is pricked with a pin, working from
the region of the lowest dermatome upwards, and the level at
which general arousal occurs is noted. It is necessary to exclude
spinal reflex movements which occur below the level of the
lesion and some experience is required in distinguishing these
from movements for which the integrity of higher pathways is
necessary. In the 79 survivors* the initial assessment proved to
be accurate to within two segments on the trunk and one segment
in the lower limbs. It was even possible to detect asymmetrical
and partial loss. The value ofneurological experience in carrying
out this test is of greatimportance, fortheassessment donebythe
junior staff was on occasion widely out. There does seem a
general tendency for the sensory level to shift one dermatome
caudally over the first year to 18 months. It is impossible to be
certain whether this observation indicates true neurological
improvement, which could conceivably occur either by ingrowth
of cutaneous nerve fibres from the segment above into the
denervated area below or by restoration of root function. Over
the trunk, owing to the small size of a neonate, precision to
within one or two segments cannot be expected.
Three groups were defined as (a) high sensory level T5-T10,
(b) intermediate sensory level Ti1-L3, and (c) low sensory level
L4 and below. (This last group contained four children in
whom no sensory loss was detected.) The overall disability in
cases with low sensory level was significantly better than cases
with high sensory level (P<0-001) (table I).
TABLE I-Overall Disability Related to Sensory Level in 79 Cases of Myelo-
(n = 32)
(n = 12)
High (n = 33)
Intermediate(n -28) ^
Low (n = 18)
No sensory loss
Overall disability of cases with low sensory level significantly better than cases with
high sensory level, P <0-001.
The importance of the sensory level as a predictor has not
been previously stressed because attention has been focused
more on motor activity of legs and on external or the radio-
logicalt extent of the lesion. The upper limit of neither of these
*In two babies the sensory levels or the two sides of the body were widely
different so that it was not possible to give a single index in either case.
These two children (1 alive and 1 dead) have therefore been excluded from
the analyses based on sensory levels, so that in the tables concerned with
sensory level (tables I-VI) the numbers ofsurviving cases is 79 and deaths 32,
instead of 80 and 33 respectively.
tThe radiological extent of the myelomeningocele
vertebrae in which the interpedicular distance is at least 2 mm greater than
normal, taking into account the region of the spine, the interpedicular
distance of the vertebrae above and below the lesion, and any abnormalities
of the vertebral arch seen on the lateral film. In several children there were
areas where the vertebrae showed increased interpedicular distance at sites
removed from the myelomeningocele. The number of dysraphic vertebrae
ranged from 0 to 17 (excluding sacrum). There was no significant correlation
between the number of dysraphic vertebrae and the final handicap.
is defined as those
BRITISH MEDICAL JOURNAL
27 OCTOBER 1973
last two necessarily coincides with the sensory level, and the
disparityinmany of our cases was wide. Of the cases in this
series 66% had asensory level above the upper extent of the
radiological dysraphism, the most extreme being seven derma-
tomesabove; in16% they coincided; in 18% the sensory level
was below the upper level of the dysraphism, in one case nine
dermatomes below. In most of the cases withhigh or extensive
lesions thesensorylevel is above theradiological lesion; in the
lower or smaller ones it is more often at or below the upper
radiologicallimit of the lesion.
In the five cases whosedisability was minimal (see fig.), and
in whom the sensory levels were at L5 or below (table I), the
levels of the external lesions were strikingly varied, being mid-
thoracic in one, lumbosacral in two, and sacral in two. In all
five the sensory level was below the upper limit of the radio-
In this seriesonly open myelocelesareconsidered; there are no
cases in which the back lesion was completely skin covered.
The cases in this series are classified as myeloceles because all
the lesions contained neural elements and ineverycase the cord
wasabnormallysituated. Most of the lesions were more than
5 cm inlengthand in 59 the abnormal cord or caudaequinawas
presentingon the surface of the lesion as aplaque which was
plainlyvisible beforeoperation.In 21 cases aplaquewas not so
easily visible though its presence always manifested itself by
some dimpling of the overlying membrane, an area of granul-
ation, or an area of scar tissue. Where there was not an obvious
plaque on the surface the outcome seemed to be more favour-
able, and in this group were all the five children who were
finally left with minimal disability only.
Cases of severe and extensive paralysis in whom there was no
voluntary activity at all in the legs except possibly some hip
flexion were compared with cases in whom the paralysis was
Results of treatment in 80 children with myelomeningocele.
TABLE Ii-Outcome Related to Predictive Features
Disability of Survivors (n = 80)*
(n = 33)
Disability of Survivors (n = 80)*
(n = 33)
(n = 5) (n =32) (n =31) (n = 12)
(n = 5)
(n = 32) (n= 31) (n = 12)
Lowsensorylevel L4 +
below(n = 18)
High sensory level T1O
+ above(n = 33)
level T1 1-L3 (n = 28)
(n = 21)
Combined low sensory
level and no plaque
(n = 9)
(n = 37)
Head size <90th centile
at birth(n = 60)
Sacral(n = 5)
No infection (n = 73)
Neuralplaque(n = 59)
Combined high neutral
plaque (n = 30)
Head size > 90th centile
at birth (n = 20)
(n = 26)
sacral (n = 10)
19 1017 P<0-02
Infection (n = 7)
*In data on sensorylevel 79 cases and 32 dead(two omitted on account of widely asymmetrical sensory level).
tRelative numbers of cases with minimum or moderate disability compared with numbers with severe orveryseveredisability.
TABLEiII-Mobility, Intelligence, Continence,and OverallDisability Related to Sensory Level in 79 Children* Operated on for Myelomeningocele
High, T5-T1O (n = 33)
Intermediate,Ti 1-L3 (n = 28)
Low, L4 or lower (n = 18)t. .
*One case omitted on account ofwidely asymmetrical sensory level.
tlncludingfour children with no sensory loss.
Goodmobility = able to walk at least 20yards (18 metres) with or without aids or in the case of young children, able to bear weight on legs; poor mobility = walking only a
fewstepsorhavingsome function inlegsifwalkinghas notyetbeenachieved; no mobility = no useful walking and severe paraplegia.
In children over 5yearsaboveaverage intelligence = I.Q. >110; average = I.Q. 110-90; below average = I.Q. 90-70; severely retarded = I.Q. <70; for children under 5
Continence refersonlyto children over 3years (n = 61) andimplies the social rather than neurological condition.
Lowsensorylevelsignificantlybetter thanhighformobility (P <0001), continence (P <0-05), overall disability (P <0-001). Low sensory level significantly better than com-
bined intermediate andhighforintelligence (P<005).
. -I- - -
confined to a lower level and in whom the quadriceps in one or
both legs showed some voluntary activity. Those with active
quadriceps showed less overall disability but this relation was
not as well defined as that between sensory level and overall
disability. Errors occurred in this series as a result of poor
condition of the baby, from reflex movement, or after breech
delivery. Sixteen of the 80 survivors had been delivered by
breech, and it was noticeable that many of them showed better
mobility when later assessed than would have been expected
from examination at birth. Eventual mobility in the whole
series was predicted slightly better by sensory level than by
motor function at birth (the tau coefficient of rank correlation
was 0 45 for sensory level and final mobility as opposed to 0*42
for movement at birth and final mobility).
Hydrocephalus present at birth was defined as having a head
circumference greater than the 90th centile using O'Neill's
(1961) norms based on head circumference related to weight.
The birth weights of the babies in this series range from 1,500 g
to 4,200 g. Children born with hydrocephalus subsequently
showed a somewhat greater overall disability and a lower range
of intelligence than the rest ofthe series. They did not, however,
show any difference in mortality (table II).
Despite the good correlation between height of sensory level
and overall disability (table I), that between extemal level
and overall disability was poor. We agree with others (Lorber,
1972) that there are a higher proportion ofseverely handicapped
cases among those presenting with thoracolumbar and thoraco-
lumbosacral lesions than in those with thoracic,
lumbosacral, or sacral lesions. But we do not find this feature
of high predictive value.
The effects of infection-meningitis, ventriculitis, and septi-
caemia-are added to table II despitethe fact that this complica-
tion could rarely have been anticipated at birth. In our series
the final results after treating these cases were uniformly bad.
The onset of this complication should perhaps be a contra-
indication to further treatment.
PREDICTIVE VALUE OF SENSORY LEVEL
The good correlation between sensory level and overall disability
(P<0 001) detailed in table I led to an analysis of the three
component parameters-mobility, intelligence, and continence.
The results are su.mmarzed in table III which shows that
mobility, intelligence, and continence were
better in the low sensory level group. These results are in
agreement with the findings of Smith and Smith (1973). From
table III it can be seen that most of the children whose sensory
level was high (T5-T10) had poor mobility, low intelligence,
and were incontinent. Most were severely disabled, and there
were none whose disability was minimal. By contrast, the
children with a low sensory level (L4 or below) mostly had
good mobility, all were of near average intelligence, and many
were continent. In this group there were no severely retarded or
chairbound children, no cases of severe kyphosis, and no blind
or partially blind children, and none who had suffered cold
injury. All the five children with minimal disability were in this
group, their sensory levels were L5 or below (two of them had
no sensory loss). Three of these five children had had their
hydrocephalus treated by ventriculoatrial shunt.
BRITISH MEDICAL JOURNAL
27 OCTOBER 1973
The incidence of hydrocephalus was significantly higher
(P<0 05) among cases with high sensory level than those with
low sensory level. Of the cases with high sensory level 36%
were born with head circumference greater than the 90th centile
compared with 2% of cases with low sensory level. Altogether
73 of the 80 survivors (91%) required reliefofhydrocephalus by
ventriculoatrial shunt. The following complications were also
significantly higher among cases with a high sensory than with a
low sensory level: eye defects (P<0.05),* kyphosis (P<0 001),
chilblains (P<0 05), and faecal incontinence (P<0 05). There
was no significant correlation between height of sensory level
and the incidence of ventriculitis, convulsions,t urinary in-
fections, fractures, or pressure sores.
Lorber (1972) has given certain signs apparent at birth which
imply severe disability and are adverse criteria to selection for
treatment-namely (1) thoracolumbar or thoracolumbosacral
lesions; (2) severe paralysis; (3) kyphosis; (4) severe clinical
hydrocephalus; (5) other gross congenital defects or cerebral
birth injury. The first four of these correlated in our series with
high and intermediate sensory level (see Appendix). They are
useful parameters where sensory level is difficult to define.
The most unexpected and probably one of the most important
relations found from the data was that between deaths from
renal failure and sensory level. Reference to table II shows that
in treated cases of myelomeningocele mortality from all causes
bore little relation to initially unfavourable features, except in
cases of infection or in the presence of a neural plaque. Thus
many severely disabled patients continue to live while some of
the best cases with low sensory level succumb. There was,
however, in this series a clear correlation between deaths
from renal causes and sensory level. Here again our findings are
remarkably similar to Lorber's. In his series of 200 cases renal
deaths occurred only in cases with adverse criteria for selection;
in our series of 113 they occurred only in cases with high or
intermediate sensory levels (table IV). More specifically all
seven renal deaths in our series occurred in the 54 cases with a
sensory level of Tl l or above (significant, P<0 01). Since this
TABLE iv-Renal Deaths in Lorber's (1972) Series and Present Series
No. of Cases
Lorber's series (200 cases):
With adverse criteria
No adverse criteria .49
Present series (111 cases):
High and intermediate sensory levels
Low sensory levels
review was made two more children are known to have died
from renal causes and their sensory levels were T8 and T10.
In patients dying from renal causes there is associated a
uniformity of high sensory level despite a diversity of both
extemal and radiological levels as shown in table V. A sensory
level of TI 1 or above denotes an interruption in the cord above
the sympathetic outflow to the kidney (T12-L1) and ureters
(Li + L2) (Hamilton, 1966). There is evidence that when
interruption of the spinal cord occurs above or at the level of
the sympathetic outflow to the legs (Ti1-L2) the sympathetic
nerves deprived of control from higher centres give rise to
abnormal and exaggerated vasomotor
(Porter, 1967). In our series we found chilblains in the legs
were associated with sensory levels from T6 to L3 but none
occurred when the sensory level was below L3. It is tempting to
speculate that similar sympathetic overaction may occur in the
upper renal tract in association with cord lesions above or at the
renal sympathetic outflow giving rise to abnormalities
in the legs
*Visual defects consisted chiefly of squints, but two children were totally
blind. Three further children had loss of vision in one eye due to chronic
comeal ulceration associated with loss of corneal sensation.
tEpilepsy occurred in 15 of the 73 children with ventriculoatrial shunts but
in none of the seven children without shunts.
BRITISH MEDICAL JOURNAL
27 OCTOBER 1973
vasomotor control or neuromuscular mechanisms in the upper
renal tract. This may explain the increased renal mortality
associated with cord lesions at or above Tll.
TABLE V-External, Radiological, and Sensory Levels in Children Dying of
T12 - sacrum
T4-T12, L2 - sacrum
T7 - sacrum
Tll - sacrum
It is clear that in deciding whether or not to advise surgery
no single criterion is going to suffice. The age and competence
of the parents and their wishes in the matter, the number of
children in the family, and the availability of specialist medical
services and appropriate educational and training facilities must
all play a part in the decision. But the fundamental question of
the possibility of achieving a reasonable degree of independence
is governed almost totally by the neurological deficit. In assessing
this we wish to emphasize the importance of the sensory level.
We would like to thank the paediatricians in the East Anglian Region
as well as many other paediatricians, medical officers of health, and
general practitioners for their help in providing much of the in-
formation on which this report is based. We are also indebted to the
many school-teachers, psychologists, and physiotherapists in the
region who gave invaluable help with the disability assessments. Our
thanks are due to Dr. E. C. Poulton, ofthe Medical Research Council,
for doing the statistics, to Mrs. Sheldon for the figures, and to Mr.
A. E. Holmes and Mr. A. H. G. Murley for advice in the preparation
of this paper.
This research was carried out with financial support from the
East Anglian Regional Hospital Board and The United Cambridge
APPENDix-Lorber's (1972) Adverse Criteria Related to Sensory Level in 79
(n = 33)
(n = 28)
(n = 18)
lesions (n = 35)
Grade4or5paralysis(n = 49)
Kyphosis (n = 20)
Severe clinical hydrocephalus* (n = 5)
*Severe clinical hydrocephalus as defined by Lorber indicates a head circumference
at least 2 cm greater than the 90th centile using O'Neill's norms.
Brocklehurst, G., Gleave, J. R. W., and Lewin, W. S. (1967). British Medical
Journal, 1, 666.
Egan, Dorothy F., Illingworth, R. S., and MacKeith, R. C. (1969). Develop-
mental Screening 0-5 years. London, Heinemann.
Hamilton, W. J. (1966). Textbook of Anatomy. London, Macmillan.
Hide, D. W., Parry Williams, H., and Ellis, H. L. (1972). Developmental
Medicine and Child Neurology, 14, 304.
Lorber, J. (1971) Developmental Medicine and Child Neurology, 13, 279.
Lorber, J. (1972). Archives of Disease in Childhood, 47, 854.
O'Neill, Eileen M. (1961). Archives of Disease in Childhood, 36, 241.
Porter, R. W. (1967). Developmental Medicine and Child Neurology, Suppl.
No. 15, p. 62.
Sheridan, Mary D. (1960). Developmental Progress in Infants and Young
Children. London, H.M.S.O.
Smith, E. Durham, and Smith, G. Keys. (1973). To be published.
Stark, G. D., and Drummond, Margaret, G. (1973). Archives of Disease in
Childhood. In press.
Early Results of Selective Treatment of Spina Bifida Cystica
British Medical Journal, 1973, 4, 201-204
The results of a policy of "selection" in treating infants with
spina bifida cystica over 21 months are presented. Of 37 new-
born infants referred on the first day of life 25 were not
treated because of the severity of their condition, as defined
by exact criteria. All died under 9 months of age. In con-
trast, only one treated infant died and the rest are either
normal or moderately handicapped. Selection for treatment
is offered as the best but not a good solution to an insoluble
Two major revolutions have occurred in the past 15 years in
Department of Child Health, University of Sheffield, Sheffield
J. LORBER, M.D., F.R.C.P., Reader in Child Health
the treatment of myelomeningocele. The first was the enthusi-
asm or moral compulsion to treat all infants, irrespective of
the degree of their handicap-largely the result of the insis-
tence of the Sheffield team (Sharrard et al., 1963; Zachary,
1968). The main reason for this enthusiasm was the introduc-
tion of the ventriculoatrial shunt in 1958. This procedure
was able to control hydrocephalus effectively for the first time.
use of unidirectional valve systems to control the associated
hydrocephalus, led to hopes that were unfortunately not ful-
filled. Analysis of the results (Lorber, 1971; 1972 a) have
shown that treating all infants still resulted in a high mortal-
ity rate in those cases severely affected at birth and yet led to
the prolonged survival of many severely handicapped children,
with gross paralysis, multiple deformities of the legs, frac-
tures, kyphosis, scoliosis, incontinence of urine and faeces with
frequent secondary effects of hydronephrosis, chronic pyelone-
phritis, and arterial hypertension. Hydrocephalus was usually
well controlled with shunt therapy-in those who needed
operation-but the complications of shunt therapy are ex-
tremely oommon, requiuing repeated operations. The mor-
tality rate from these complications alone was 20% within
seven years of the first shunt operation in a large group of