Personal Practice

Archives of Disease in Childhood, 1971, 46, 539.

Neonatal Assessment of the Child with aMyelomeningocele

GORDON D. STARKFrom the Royal Hospital for Sick Children, and University of Edinburgh

Before control of hydrocephalus was feasible,surgical closure of a myelomeningocele was seldomattempted, and detailed neurological assessmentattracted little or no interest. However, with theadvances in neonatal surgery, infants who wouldpreviously have been registered as 'stillborn' arenow reaching children's hospitals which mustincreasingly face the ethical problems they present.With traditional Scottish scepticism the RoyalHospital for Sick Children in Edinburgh is amonga minority of units in which it is felt that some ofthese infants (at present 40-45%) are so severelyhandicapped and so unlikely to survive thatoperation cannot be justified. In most largecentres in the United Kingdom, on the other hand,it is standard practice for every infant with amyelomeningocele to be subjected to early operationirrespective of his condition and prospects (Nash,1970; Mawdsley and Rickham, 1969; Zachary andSharrard, 1967; Forrest, 1967). There is, however,growing disillusionment with the results of routineearly closure among former enthusiasts (Lorber,1971).Whatever the ethical position regarding operation,

few would disagree that every infant born with spinabifida deserves an opportunity for full assessmentsoon after birth. Such assessment provides thebasis for prognosis and, if operation is not routine,selection in each case. If the lesion is repaired,this assessment is invaluable as a baseline for futureassessment and in planning of orthopaedic, uro-logical, and other aspects of management. Further-more, neonatal examination is of considerableresearch interest since it is only at this stage thatthe effects of the lesion can be studied relativelyunmodified by such postnatal factors as infectionand surgical intervention.The necessary expertise and facilities for assess-

ment and management are best provided in regionalpaediatric centres with a particular interest in the

problem and able to assemble a multidisciplinarymedical and surgical team backed up by a com-prehensive laboratory service. General practitionersand the staff of every maternity hospital shouldknow the location of their regional centre and beprepared to transfer affected infants as soon aspossible after birth. During transport, the infantmust be kept warm in a portable incubator, thelesion protected by a dressing of sterile gauzesoaked in normal saline, and accompanied by aspecimen of maternal blood for cross-matching iifnecessary.The outline which follows is based on practice in

the neurological unit of the Royal Hospital for SickChildren in Edinburgh which serves a population ofabout 1 million and admits 35 new cases a year.More than 70% are admitted under 6 hours andmore than 95% under 12 hours.

History TakingFrom the child's father details are obtained of the

family and social history. The family history maybe supplemented later at the time of genetic coun-selling and the medical social worker's early contactwith the father is later followed by more completeevaluation of the family's needs and resources.The family doctor or Maternity Hospital suppliesthe obstetric historywhichis frequentlycomplicated:of 130 consecutive infants, only 66% were born byspontaneous or low forceps delivery, 14% requiredintubation at birth, and birth injury was frequent(Stark and Drummond, 1970).

Clinical ExaminationThe initial assessment is carried out by a paediatric

neurologist joined later by surgical and orthopaediccolleagues. Infants who are cold and apathetic(one-third have rectal temperature below 35 5 °Con admission) or in otherwise poor condition arerewarmed before examination. During examina-

539

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Gordon D. Starktion the child remains in an incubator such as theIsolette model C 86 which provides good access.

General examinationParticular attention is paid to the infant's general

state (temperature, colour, respiration, level ofactivity) which can profoundly influence theneurological findings and fitness for operation.Since congenital malformations seldom come singly,systematic examination is carried out in every caseand has led to the finding of such diverse lesions asimperforate anus, cleft palate and congenital heartdisease.

Examination of spinal lesionThe type of lesion is first noted. The simple

meningocele and closed myelomeningocele in whichthere is no exposed neural plate must bedifferentiated from the commoner and more seriousopen myelomeningocele with which this discussionis concerned. The whole lesion and the exposedneural plate are measured and the vertebral level andextent noted. Though it is generally reported thatlumbar and lumbosacral lesions predominate, in ourseries 66% have been thoracolumbar or thoraco-lumbosacral, 16% lumbar, and 18% lumbosacral.In general, the larger the neural plate and the higherits situation, the greater the neurological deficit.Moreover, thoracolumbar lesions are almost invari-ably associated with hydrocephalus. Recognitionof asymmetrical lesions with severe spinal deformity(Fig. 1) is important since they may be associatedwith a split cord penetrated by a bony spur whichmust be removed at operation. The membranesurrounding the plaque is inspected carefully fortears which occur in 10% and predispose tomeningitis. The width of the bony defect andavailability of skin for repair are of obvious concern

to the paediatric surgeon, and the presence of spinaldeformity, which may require osteotomy, to theorthopaedic surgeon.

Neurological assessmentCerebral function. Time is too limited for a

standardized examination of the type described byPrechtl and Beintema (1964), which is, in any case,not readily applied to the infant with a spinal lesion.Attention is, however paid to the infant's state ofalertness and ability to fixate objects with his eyes.Signs of hydrocephalus are sought. An occipito-frontal circumference above the 90th centile isindicative of severe hydrocephalus: in Lorber'sseries (1961) the cerebral mantle was less than 25mmin every such case. However, equally severehydrocephalus may be present in the 70 to 80% ofinfants with less conspicuous head enlargement atbirth. In them, separation of the sutures, especiallylambdoidal, tense fontanelle, distended scalp veins,and papilloedema (which does occur) are more usefulsigns. The asymmetrical tonic neck reflex is verystrong if ventricular pressure is high and may returnafter the age of 6 months in the child with a blockedcerebrospinal fluid shunt.The presence of hydrocephalus can, however, be

assumed in 85 to 90% of patients and its degreerequires investigation. Cranial nerve lesions are notuncommon: apart from sixth nerve lesions fromraised intracranial pressure, fasciculation of thetongue and stridor may occur. Such lower cranialnerve disorders may be related to brainstemcompression from the Arnold-Chiari malformation.

Spinal cord function(i) Lower limbs. The rare cervical myelo-

meningocele may interfere with the upper limbsbut, for practical purposes, assessment of residual

FIG. 1.-Hemimyelomeningocele with typical spinal deformity.

540

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Neonatal Assessment of the Child with a Myelomeningocele

L1 L2 L3_ L4 L5 S S2 S3ILIO PSOAS

SARTORIUSPECTINEUSGRACILSISADD. LONGUS

ADD. BREVISADDUCTOR MAGNUSQUADRICEPS

OBT. EXT.TIB. ANT.

TIB. POST.TEN. FAS. LATAGLUT. MED. & MIN.SEMIMEMBRANOSUS

LSEMITENDINOSUSEXT. HALL. L.1

EXT. DIG. L.PER. TERT.PER. BREVISPER. LONGUSHPLAT. ROT.GASTROCN.SOLEUS & PLANTBICEPS FEMORISGLUTEUS MAX.

FLEX.HALL.L1-4FLEX.DIG. L. 9.

________| FOOT INTRINSICSFIG. 2.-Summary of Sharrard's findings (Sharrard, 1964a).

spinal cord function depends on examination of the 1971a) andtrunk and lower limbs. There are two prerequisites Type Ifor this examination. normal dovThe first is familiaritywiththe segmentalinnervation completely

of lower limb muscles on which subject the standard loss of sensi

anatomical texts are singularly unhelpful. By not so muccontrast, the classical paper of Sharrard (1964a) is spinal shoc]invaluable and shows that lower limb muscles are later recoveinnervated for the most part from one or two spinal of lesion, tdcord segments. Sharrard's findings are summarized the lower liin Fig. 2. cord invol'The second requirement is awareness of the abdominal

different patterns of neurological abnormality which and the legmay be encountered (Stark and Baker, 1967; Stark, With presei

FIG. 3.-Type I lesion.

which will now be outlined.(Fig. 3): In some infants, function iswn to a certain level below which it isabsent, resulting in flaccid paralysis

ation and reflexes. This is probably duech to failure of cord development as to:k from birth trauma, since a proportioner distal reflex activity. With this typehe pattern of weakness and deformity inimbs depends on the upper level of spinalvement. If this is T8, the paralysedmuscles bulge paradoxically on crying

gs are flaccid but undeformed (Fig. 4).rvation of function to L4, the legs show

FIG. 4.-Flaccid paraylsis below T8.

541

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Gordon D. Stark

FIG. 7.-Foot deformity with paralysis below Sl.

be troublesome (Fig. 7): in this infant both dorsi-flexors (L4-5) and calf muscles (S 1-2) are strong,but paralysis of the intrinsic muscles of the foot(S2-3) has allowed a break to occur at the midtarsaljoint producing a rockerbottom or boat-shaped foot.In such infants, with involvement of the lower sacralsegments, inactivity of pelvic floor muscles (S3--4)

FIG. 5.-Flaccid paralysis below L4. results in a flat-bottomed appearance (Fig. 8), withabsence of the natal cleft and a patulous anus.

characteristic deformities illustrated in Fig. 5.Psoas, adductors, quadriceps, and tibialis anteriorinnervated from L1-4 are strong but their anta-gonists innervated from L5 and below are paralysed.Fig. 6 shows a child with normal function down toL5: toe extensors, peronei, and medial hamstrings ;(L5-S1) are active, but calves (S1-2) are very weak.With good function down to S1, foot deformity may.

FIG. 8. Flaccid pelvic floor.

Type II (Fig. 9). (a) In others, one finds a 'gap'in cord function with loss of motor, sensory, andreflex activity but preservation distally of intactbut isolated cord segments. In the latter territory,spasticity and stretch reflexes may be striking, e.g.ankle and toe clonus. Exteroceptive tonic reflexes(Vlach, 1968) can be elicited, e.g. stroking of thedorsum of the foot evokes toe extension, stimulation

-i : _ E ~of the lateral aspect of the leg, eversion of the foot,FIG. 6. Flaccid paralysis below L5. and of the lateral aspect of the thigh abduction of

542

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Neonatal Assessment of the Child with a Myelomeningocele- M_i _ - _ ' %'

FIG. 9.-'a) Type IIa lesion. (b) Type IIb lesion.(c) Type IIc lesion.

the hip. More complex reflex activity may followstimulation of the perineum and perianal region: inaddition to contraction of the anal sphincter, thereis commonly irradiation of the response to severalmuscles of sacral innervation. Fig. 10 shows sucha child with flaccid paralysis of muscles innervatedfrom lumbar segments but spasticity of musclessupplied by S1-5 including biceps femoris, calf, and

FIG. 10.-Type IIa lesion showing spastic deformities.

FIG. 11.-Spastic pelvic floor.

short toe flexors. In such infants who have spas-ticity of the pelvic floor the anus is hidden in a deepnatal cleft (Fig. 11).

(b) In others still the 'gap' is narrow, evenimperceptible, and the lesion amounts to spinalcord transection. Though there is no movementof the legs on vigorous crying and no centralresponse to pin-prick stimulation, the slighteststimulation at any point on their surface will usuallyevoke an uninhibited flexion withdrawal reflex(Fig. 12).

(c) The transection may be incomplete so that thechild has a spastic paraplegia but incomplete loss ofvoluntary movement and sensation.

The relative frequency of the above neurologicalpatterns in a series of 100 consecutive infantsexamined within 24 hours of birth (the majoritywithin 6 hours) is shown in the Table. It is evidentthat though myelomeningocele is often quoted as atypical lower motor neurone lesion, this is not soand upper motor neurone lesions predominate.These figures are, in fact, an underestimate of the

FIG. 12.- Type IIb lesion. (a) inert, undeformed legs.(b) Flexion withdrawal reflex.

543

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Gordon D. StarkTABLE

Neurological Patterns in 100 Consecutive InfantsExamined under 24 hours

Normal .8%Type I lesion .28%Type II lesion .64%- Ila :37%

i Ilb :18%LIIc: 9%

latter as infants with normal function at birth tendto develop upper motor neurone lesions rapidly andthose initially classified as Type I may later recoverdistal reflex function. A high incidence of uppermotor neurone lesions has also been found byGuthkelch (1964) and Brocklehurst, Gleave, andLewin (1967).As already noted, in about 5% of cases, the cord

is split and only one-half is involved in the myelo-meningocele (Fig. 13). While one leg is more orless normal, the other is affected in one of the waysalready described (Duckworth et al., 1968; Stark,1968; Ericsson et al., 1970). Nerve root lesions asillustrated diagrammatically in Fig. 14 are extremely

FIG. 13.-Split cord with hemimyelomeningocele.

rare. In such patients normal function is founddistal to a zone of flaccid paralysis and sensoryloss.

Sensory loss in the trunk and lower limbs usuallyfollows motor loss within one or two segments.Vasomotor control is commonly disturbed: the skinof the legs may be cold and mottled or show adiffuse flush below the level of the lesion. Apartfrom Porter's (1968) study, these vasomotorphenomena have received scant attention and theirclinical significance in relation to later chilblainsand trophic changes in the legs is uncertain.

FIG. 14.-Nerve root lesion.

Technique of examination of the lower limbs willnot be considered in detail, but a few points areworthy of note. Sensory testing is best carried outwith the infant quiet, but if he is very lethargic,examination is worthless and must be repeated.The first aim is to determine the lowest level ofnormal sensation: starting in the lowest sacralterritory, i.e. perianal region, the skin is stimulatedover the posterior aspect of the buttocks, thighs,and legs, and then upwards over successive derma-tomes of the anterior surface and on to theabdomen. All the while the baby is watchedclosely for a facial grimace, a cry, or a Mororesponse. Having detected the sensory level, anote is made of any territory from which only areflex response can be obtained. Motor testing,which is performed next, requires that the infantshould be warm, hungry, and active showingvigorous arm movements. Care must be taken toavoid pressure on the neural plate which may evokebrisk leg movements which can be mistaken fornormal.

Initially 'voluntary' or rather spontaneousmovement is assessed, and for this purpose it ispermissible to activate the child by stimulation ofthe upper limbs. By appropriate positioning, eachmuscle can be made to operate with gravityeliminated or against gravity and, by palpation, thestrength of contraction can be graded, e.g. on theM.R.C. scale (Medical Research Council, 1942).Some muscles, e.g. glutei, are more difficult to assessthan others, but with patience and practice it ispossible to assign a motor level to each side, i.e. thelowest segmental level of voluntary motor function.

544

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Neonatal Assessment of the Child with a MyelomeningoceleThen-and only then-are the lower limbs andperineum stimulated directly (a straightened paperclip or small screwdriver are useful instruments) andthe extent of purely reflex movement noted. Anydifficulty in distinguishing voluntary from reflexmovement can usually be dispelled by observationof the lower limb response to Moro, asymmetricaltonic neck, and stepping reflexes. A normalresponse indicates integrity of long spinal tracts.The M.R.C. scores are recorded on a chart ofmuscles arranged in order of segmental innervation,purely reflex activity being indicated by the letter'R'. The lower limb findings can be summarizedby recording for each side:Motor level: Motor isolated cord:Sensory level: Sensory isolated cord:

(ii) Bladder and bowel. Clinical and neurologicalexamination is of some value in the prediction ofincontinence (Stark, 1971b). Frequent, smallvolume dribbling of urine (Fig. 6), increased bycrying, movement, or suprapubic pressure is awarning of future incontinence, whereas observationof periodic micturition with a good stream is moreoptimistic. Constant leakage of meconium from apatulous anus similarly presages later problems butbowel training is more readily accomplished(Forsythe and Kinley, 1970). There is fairlyreliable correlation between the neurological picturein the lower limbs and bladder function (Stark,1968). If even one leg is normal or normal apartfrom a mild upper motor neurone lesion, normalbladder function can be expected. Infants whohave neither voluntary nor reflex function in musclesinnervated from S2-4 (lateral hamstrings, calf,intrinsics, and anal sphincter) tend to have completebladder paralysis. Those with either partialvoluntary function or purely reflex activity in S2-4usually have an active detrusor but an efficient auto-matic reflex bladder is rare and outlet obstructioncommon (Stark, 1969a).

Lower limb deformityLower limb deformities in myelomeningocele are

not 'associated malformations' but result frommuscle imbalance, i.e. imbalance between normaland flaccid muscles in Type I lesions; betweenspastic and normal or flaccid muscles in Type IIlesions. The range of passive movement at thelower-limb joints is recorded and deformitiesdescribed in detail by the orthopaedic surgeon.An accurate 'spot diagnosis' of the segmental

motor level can often be made from a glance at theposture and deformity of the lower limbs. This is,however, unreliable, as fixed deformity may be due

FIG. 15.-Congenital deformities in paralysed legs.

to prenatal imbalance of muscles which havesubsequently lost all function. For example, theinfant shown in Fig. 15 has total lower limb paralysisbut fixed deformities. We have seen footdeformities consistent with an L4 segmental levelin a 16-week fetus with a lumbosacral myelo-meningocele (Fig. 16). In such instances, theprenatal neurological disorder is, as it were,'fossilized' in the deformity it has produced.

a 4l

I.

FIG. 16.-Paralytic deformities in 16-week fetus withlumbosacral myelomeningocele.

545

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Gordon D. StarkConversely, if no deformity is present at birth,

future deformities can be predicted from theneurological findings in the lower limbs. Sharrard(1964b) has, for example, clearly shown how thetendency to dislocation of the hips depends on thebalance between hip flexors and adductors (L1-3)on the one hand, and the extensors and abductors(L5-S2) on the other. If imbalance is maximal,i.e. L4 motor level, the hips are likely to be dislocatedat birth or within the first year (as in Fig. 5). Withpartial imbalance, e.g. L2 or L5 motor level,increasing flexion-adduction contracture and latersubluxation can be expected, but early dislocation isunlikely. Similarly in Type II lesions, the findingof increased reflex activity in S1-5 segments suggestsprogressive equinus and knee flexion contracturesfrom spastic calf and lateral hamstring muscles.

Special investigationsThe following procedures are carried out

routinely.(a) Bacteriological swabs from the myelomeningo-

cele and the umbilicus.(b) Photography of the spinal lesion and lower

limbs.(c) Radiological examination of skull, spine, and

hips. Unsuspected vertebral anomalies are com-monly found, and by taping wire markers round themyelomeningocele its vertebral extent can bedetermined accurately (Fig. 17).

(d) Concentric needle electrode electromyography iscarried out after clinical assessment which it cannotreplace. It may reveal activity in muscles whichare difficult to assess clinically, and though essen-tially a tool for study of the lower motor neurone,it can help to differentiate voluntary from purelyreflex activity. The ear can readily detect thesynchronization between the infant's crying andsignals from the loud speaker of the electromyo-graph which suggests voluntary movement, whereasin purely reflex activity electromyographic activityis unrelated to crying. Faradic stimulation oflower limb muscles has been advocated by Stoyle(1966) but it is less precise and subject to greatertechnical limitations in the newborn. Directstimulation of the neural plate is of research interestbut not indicated as a routine in view of the dangerof infection and trauma.

(e) Air ventriculography is carried out after closureof the myelomeningocele unless signs of hydro-cephalus are absent. Lumbar air encephalography,usually a more informative procedure, is seldompracticable in these infants but has its advocates(Andersson et al., 1968). At air ventriculography,the ventricular pressure and cerebral mantle at the

FIG. 17.-X-ray of spine showing myelomeningocelemarkers.

vertex are measured, but the precise site of obstruc-tion cannot always be identified. The value ofroutine air ventriculography is not so much indetection of infants who require cerebrospinal fluidshunts as in identifying those who may well dowithout: Lorber (1969) suggests that if the cerebralmantle at the vertex exceeds 15-20 mm, an expectantapproach is justifiable, whereas a cerebral mantle ofbelow 15 mm or ventricular pressure of over 300mmH20 are indications for early shunting. Echo-encephalography is a promising technique forneonatal assessment of ventricular size (Sjogren,1965; West, 1967; Grumme and Tilley, 1970), butits reliability has been questioned by Emery (1967)and it is, at present, used in parallel with airstudies.

(f) Urinary tract investigations: as early detectionof the unsafe or obstructed bladder is vital, thoroughurinary tract assessment must be carried out inearly infancy. For technical reasons and tominimize the initial hospital stay, we now admit thechild for a few days at the age of 3 months for directcystometry, suprapubic cystography, sphincterelectromyography, intravenous pyelography, andisotope renography. Subsequent management of

546

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

Neonatal Assessment of the Child with a Myelomeningocele 547the urinary tract problem is based on the informationso obtained (Stark, 1969b).

ConclusionsEckstein (1968), among others, has cast doubt on

the predictive value of neonatal assessment, statingthat 'it is virtually impossible to assess the potentialof the newborn infant with a myelomeningocele sothat selection for surgical treatment is usuallyimpossible'. It is true that without a crystal ball,we cannot tell which child's burden of disability willlater be increased by ventriculitis or serious valvecomplications. There is, however, mounting evi-dence that neonatal assessment can indicate theminimum disability which can be expected in anewborn infant with an open myelomeningocele.Miracles may happen but they are rare in this field.

Brocklehurst et al. (1967) have shown that, evenwith early operation, the initial neurological findingschange little: at the age of 3 months 19 out of 22survivors were basically unchanged. Though post-operative functional recovery has been reported(Sharrardetal., 1963; Fernandez-Serrats, Guthkelch,and Parker, 1967), even these authors agree that nomore than preservation of pre-operative function canbe depended on. In our experience, recovery ofvoluntary function is exceptional in infants withsigns of cord transection.

Convincing evidence for the practicability ofselection for surgical treatment has been presentedrecently by Lorber (1971). Drawing on theunparalleled experience in Sheffield of earlyassessment and early operation in over 500 patients,he has shown that a poor prognosis for both life andquality of survival are associated with the followingfindings at birth, the more so if they occur in com-bination: paralysis below Li (commonly due to athoracolumbar lesion), severe hydrocephalus(occipito-frontal circumference at the 90th centileor cerebral mantle below 25 mm), kyphotic spinaldeformity, and severe associated malformations orbirth injury. It is likely that the disappointingSheffield results will lead to general reappraisal ofthe policy of routine early closure. If this is so, it isto be hoped that every infant treated in the futurewill be subjected to detailedpre-operative assessmentso that the value offollow-up studies can be increasedand criteria for selection defined with greaterprecision.

Neonatal assessment is a guide to immediatemedical and surgical treatment, and lays thefoundations for planning of future management.It is, however, only the beginning of a continuingprocess of evaluation, the emphasis of which willchange as the child's problems change and which

will later be increasingly concerned with his needsfor education, employment, and social integration.

The method of assessment outlined in this articlehas evolved from a team approach to the problem.I am indebted to medical, surgical, orthopaedic, andnursing colleagues at the Royal Hospital for SickChildren for their co-operation and helpful discussion.I am grateful to Miss C. Brydone and Mr. C. Shepleyfor assistance with illustrations and to Mrs. IreneStewart who deciphered and typed the manuscript.Table I is reproduced by kind permission of Mr.

W. J. W. Sharrard and the publishers of Annals of theRoyal College of Surgeons of England.

REFERENCESAndersson, H., Bjurstam, N., Carlsson, C. A., and Rosengren, K.

(1968). Lumbar pneumoencephalography in newboms withmyelomeningocele. Developmental Medicine and Child Neuro-logy, Suppl., 15, 17.

Brocklehurst, G., Gleave, J. R. W., and Lewin, W. S. (1967).Early closure of myelomeningocele with especial reference toleg movement. Developmental Medicine and Child NeurologySuppl., 13, 51.

Duckworth, T., Sharrard, W. J. W., Lister, J., and Seymour, N.(1968). Hemimyelocele. Developmental Medicine and ChildNeurology, Suppl., 16, 69.

Eckstein, H. B. (1968). Myelomeningocele, basic considerations.Hospital Medicine, 2, 896.

Emery, J. L. (1967). The examination of the normal and hydro-cephalic infant brain using ultrasound. Developmental Medicineand Child Neurology, Suppl., 13, 87.

Ericsson, N. O., Hellstrom, B., Nergardh, A., and Rudhe, U. (1970).Unilateral neurological defect in myelomeningocoele withnormal bladder function. Acta Paediatrica Scandinavica, 59,487.

Fernandez-Serrats, A. A., Guthkelch, A. N., and Parker, S. A. (1967).The prognosis of open myelocele with a note on a trial ofLaurence's operation. Developmental Medicine and ChildNVeurology, Suppl., 13, 65.

Forrest, D. M. (1967). Modem trends in the treatment of spinabifida. Early closure in spina bifida: results and problems.Proceedings of the Royal Society of Medicine, 60, 763.

Forsythe, W. I., and Kinley, J. G. (1970). Bowel control of childrenwith spina bifida. Developmental Medicine and Child Neurology,12, 27.

Grumme, T., and Tilley, E. J. (1970). Postnatal echoencephalo-graphy in cystic dysraphia. Developmental Medicine and ChildNeurology, Suppl., 22, 78.

Guthkelch, A. N. (1964). Studies in spina bifida cystica. V.Anomalous reflexes in congenital spinal palsy. DevelopmentalMedicine and Child Neurology, 6, 264.

Lorber, J. (1961). Systematic ventriculographic studies in infantsborn with meningomyelocele and encephalocele. Archives ofDisease in Childhood, 36, 381.

Lorber, J. (1969). Ventriculo-cardiac shunts in the first week oflife. Results of a controlled trial in the treatment of hydro-cephalus in infants born with spina bifida cystica or craniumbifidum. Developmental Medicine and Child Neurology, Suppl.,20, 13.

Lorber, J. (1971). Results of treatment of myelomeningocoele.An analysis of 524 unselected cases, with special reference topossible selection for treatment. Developmental Medicine andChild Neurology, 13, 279.

Mawdsley, T., and Rickham, P. P. (1969). Further follow-up studyof early operation for open myelomeningocele. DevelopmentalMedicine and Child Neurology, Suppl., 20, 8.

Medical Research Council (1942). Aids to the Investigation ofPeripheral Nerve Injuries (War Memorandum No. 7). H.M.S.O.,London.

Nash, D. F. E. (1970). The impact of total care with specialreference to myelodysplasia. Developmental Medicine andChild Neurology, Suppl., 22, 1.

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from

548 Gordon D. StarkPorter, R. W. (1968). Vasomotor control in the lower limbs of

children with meningomyeloceles. Developmental Medicineand Child Neurology, Suppl., 15, 62.

Prechtl, H., and Beintema, D. (1964). The neurological examinationof the full-term newborn infant. Little Club Clinics in Develop-mental Medicine, 12.

Sharrard, W. J. W. (1964a). The segmental innvervation of thelower limb muscles in man. Annals of the Royal College ofSurgeons of England, 35, 106.

Sharrard, W. J. W. (1964b). Posterior iliopsoas transplantation inthe treatment of paralytic dislocation of the hip. Journal ofBone and Joint Surgery, 46B, 426.

Sharrard, W. J. W., Zachary, R. B., Lorber, J., and Bruce, A. M.(1963). A controlled trial of immediate and delayed closureof spina bifida cystica. Archives of Disease in Childhood, 38, 18.

Sjogren, I. (1965). Echo-ventriculography in infantile hydro-cephalus. Acta Neurologica Scandinavica, Suppl., 13, 13.

Stark, G. D. (1968). The pathophysiology of the bladder inmyelomeningocele and its correlation with the neurologicalpicture. Developmental Medicine and Child Neurology, Suppl.,16, 76.

Stark, G. D. (1969a). Pudendal neurectomy in management ofneurogenic bladder in myelomeningocele. Archives of Diseasein Childhood, 44, 698.

Stark, G. D. (1969b). Aspects of the neurogenic bladder inmyelomeningocoele. Paper presented at International Sympo-sium on the Disabled Child at Bruges (In the press).

Stark, G. D. (1971a). The nature and cause of paraplegia inmyelomeningocoele. Paraplegia (In the press).

Stark, G. D. (1971b). Prediction of urinary continence in myelo-meningocele. Developmental Medicine and Child Neurology,13, 388.

Stark, G. D., and Baker, G. C. W. (1967). The neurologicalinvolvement of the lower limbs in myelomeningocele. Develop-mental Medicine and Child Neurology, 9, 732.

Stark, G. D., and Drummond, M. (1970). Spina bifida as anobstetric problem. Developmental Medicine and Child Neuro-logy, Suppl., 22, 157.

Stoyle, T. F. (1966). Prognosis for paralysis in myelomeningocele.Developmental Medicine and Child Neurology, 8, 755.

Vlach, V. (1968). Some exteroceptive skin reflexes in the limbs andtrunk in newborns. Clinics in Developmental Medicine, 27, 41.

West, K. A. (1967). Correlations between ultrasonic and roentgeno-logic findings in infantile hydrocephalus. Acta PaediatricaScandinavica, 56, 27.

Zachary, R. B., and Sharrard, W. J. W. (1967). Spinal dysraphism.Post-graduate Medical_Journal, 43, 731.

Correspondence to Dr. Gordon D. Stark, Depart-ment of Child Life and Health, University ofEdinburgh,17 Hatton Place, Edinburgh EH9 lUW.

copyright. on O

ctober 19, 2020 by guest. Protected by

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.46.248.539 on 1 August 1971. D

ownloaded from