Archives of Disease in Childhood 1993; 68: 612-616

PERSONAL PRACTICE

Management of the asphyxiated full term infant

Malcolm I Levene

In Britain, approximately one full term babyper thousand dies or is severely disabled as theresult of birth asphyxia. It is arguably the mostimportant avoidable cause of permanentneurological injury affecting the maturefetus/newborn. It is generally agreed thatclinical signs of hypoxic-ischaemic encephalo-pathy (HIE) are the best markers for adiagnosis of intrapartum 'asphyxia'. Unfor-tunately the severity of HIE can only bediagnosed retrospectively after symptoms havedeveloped. Early therapeutic intervention inthe asphyxiated baby may be important tomodify cerebral injury (see below) and there-fore there remains a need to have early markersof asphyxia such as depressed Apgar scores,delay in establishing respiration, or evidence ofsignificant metabolic acidosis on samples ofcord blood.

Although there appears to have been a fall inboth the incidence of HIE and the number ofchildren disabled by this condition in recentyears,' there is little evidence that this has beendue to improvement in postnatal management.In this paper I will consider the standard man-agement of asphyxia, potentially useful newmethods for treating the asphyxiated brain,and consider methods of deciding when towithdraw care. The brain of the full term infantresponds to asphyxia in a very different mannerto that of a premature baby and this paper onlyconsiders the management of asphyxia in fullterm infants.

Rational basis of standard managementAsphyxiated infants require expert and rapidresuscitation wherever they are born. All healthcare professionals involved in the birth ofbabies must be adequately trained andretrained in resuscitation techniques. In mostcases 'bag and mask' is sufficient to maintainventilation until someone with advanced resus-citation skills arrives. Appropriate cardiovascu-lar support must be available for infants bornwith poor or absent circulation. The tablesummarises an approach to the management ofthe asphyxiated full term infant.

Academic Unit ofPaediatrics and ChildHealth, University ofLeeds, D Floor,Clarendon Wing, TheGeneral Infirmary atLeeds, Leeds LS2 9NSCorrespondence to:Professor Levene.

GLUCOSEIt has been shown that a raised blood glucoseconcentration before hypoxic-ischaemicinjury in adolescent animals results in more

extensive cerebral injury than in those withnormal or low blood glucose.2 This has had

the effect of making paediatricians verycautious in the use of intravenous glucoseduring neonatal resuscitation. More recentdata has shown conclusively that there is afundamental difference in the way theimmature and the more mature brainresponds to glucose infusion. This is probablyin the main related to the impaired rate ofglucose transport across the immatureblood-brain barrier. The immature brainappears to be protected by raised glucose con-centrations before asphyxial insult comparedwith animals that had no additional glucose.3There is conflicting data concerning the brainprotective effect of high glucose concen-trations after asphyxia. In an immatureanimal model, administration of glucoseimmediately after a period of hypoxic-ischaemic insult resulted in significant reduc-tion in cerebral infarction,4 but others haveshown that in a slightly different rat pupmodel there was significant exacerbation ofdamage in the presence of hyperglycaemiaafter injury.5

Hypoglycaemia must be avoided during andafter resuscitation of asphyxiated babies, butfaced with conflicting reports on the role ofglucose infusion after hypoxic-ischaemic injuryit is not possible at the present time to givepractical advice concerning the role of glucoseinfusion after human birth asphyxia. Thisquestion would be best answered in a doubleblind controlled clinical study.

PREVENTION OF CEREBRAL OEDEMAIt is widespread practice to anticipate cerebraloedema and manage the baby so as to reducethe possibility that this complication maydevelop. This is done in two ways. Firstly, a

Summary of management of severe birth asphyxia. See textfor details

Immediate management:1. Establish effective ventilation2. Assist circulation if necessary

Early management:1. Restrict fluids by 20% for first two days2. Monitor blood pressure and treat hypotension vigorously3. Assess respiratory effort and

(a) ventilate if baby breathing spontaneously with arterialcarbon dioxide tension >7 kPa

(b) if baby ventilated maintain arterial carbon dioxidetension at 4-5 kPa

4. If clinical signs of raised intracranial pressure givemannitol 1 g/kg over 20 minutes and repeat if necessaryevery 4-6 hours

Anticonvulsants if:1. Frequent convulsions >3 per hour2. Prolonged convulsions lasting .3 minutes

612

on 17 April 2019 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.68.5_S

pec_No.612 on 1 M

ay 1993. Dow

nloaded from

Management of the asphyxiated full term infant

regimen of fluid restriction is often institutedand secondly corticosteroids are administered.

Fluid restrictionThere have been no studies on the effect offluid restriction in infants with cerebraloedema. In general, measures to reducecerebral oedema probably have no effect onlong term neurological outcome (see below)and it is difficult to argue that routine fluidrestriction has any advantage in this respect.

Fluid restriction may be important inasphyxiated infants who have complicationssuch as inappropriate secretion of antidiuretichormone and renal compromise. Fluid reten-tion occurring as the result of these two con-ditions may further compromise the infant andfor these reasons I recommend restrictingfluids by 20% of the normal regimen for thefirst two days of life or until such time as thebaby's renal function recovers.

Hypotension is a relatively common compli-cation of birth asphyxia and may be due toreduction of the circulating blood volume andtherapeutic dehydration may exacerbate thiscondition. Plasma infusion may be required inhypotensive asphyxiated infants. Hypogly-caemia and overt dehydration as a result offluid restriction must be avoided.

CorticosteroidsThere are no data to support the use of corti-costeroids in the routine management of birthasphyxia. Studies in adults and in animalmodels have failed to show any benefit inreducing brain swelling or improving outcome.In animal models of asphyxia, corticosteroidshave shown either a detrimental effect6 or nobenefit at all.7 Corticosteroids are associatedwith a number of actual or potential sideeffects including hyperglycaemia, hyperten-sion, susceptibility to infection, gastrointestinalhaemorrhage, and restriction of later braingrowth. In my view, corticosteroids should notbe used in the management of birth asphyxia.

MANAGEMENT OF CEREBRAL OEDEMAIn very severely asphyxiated infants, raisedintracranial pressure (ICP, a sustainedincrease to 1x33 kPa (>10 mm Hg)) lasting for20 minutes or more occurs in 70% of cases.8The management of intracranial hypertensionincludes hyperventilation (lowering of thearterial carbon dioxide tension) and infusion ofosmotic agents.

HyperventilationHyperventilation lowers ICP by cerebral vaso-constriction with reduction of cerebral bloodvolume and consequent decrease in totalintracranial volume. An important componentof the pathophysiology of postasphyxial injuryis cerebral hypoperfusion and it is possible thathyperventilation may exacerbate impairedreperfusion to the brain. There have been noclinical trials of hyperventilation in the

asphyxiated newborn. Babies who havesuffered significant birth asphyxia may spon-taneously hypoventilate with resulting hyper-capnia and increased cerebral blood volumewhich is probably undesirable. For this reasonall encephalopathic babies should have anarterial carbon dioxide tension measurementand if this is >7 kPa (53 mm Hg) then thebaby should be electively ventilated. Themechanical ventilator should be adjusted tomaintain the arterial carbon dioxide tension atabout 4-5 kPa (34 mm Hg).

Osmotic agentsOsmotic agents (mannitol or glycerol) are usedto reduce cerebral oedema by increasing serumosmolality. In a neonatal animal model manni-tol significantly reduced brain water contentwhen given immediately after an hypoxic-ischaemic event,9 but it did not reduce theseverity of distribution of brain damage intreated compared with untreated animals.There have been no neonatal randomised con-trolled studies of mannitol or glycerol in themanagement of intracranial hypertension.Marchal et al in an uncontrolled study gavemannitol to 225 babies with the diagnosis ofasphyxia,1I although the precise indications fortreatment were quite varied. Early treatmentwas defined as mannitol infusion (1 g/kg)before the baby was 2 hours of age. There weresignificantly fewer deaths (p=0 005) and thesurvivors had better neurological outcome(p=0-014) in the early treatment group com-pared to those treated after 2 hours. Leveneand Evans showed that mannitol (1 g/kg over20 minutes) reduced ICP in a small number ofseverely asphyxiated infants with intracranialhypertension (>1-33 kPa or >10 mm Hg).'1There was a concomitant rise in cerebral per-fusion pressure 60 minutes after starting themannitol infusion. The effect of mannitollasted for approximately four hours.

Providing there is adequate renal function toallow excretion of mannitol, it appears to be arelatively safe agent in the management ofcerebral oedema. Its efficacy remains in doubt,but I would recommend its use in infantswith a bulging fontanelle or in whom thereis the clinical impression of intracranialhypertension.

There are very few studies to evaluate therole of intracranial pressure monitoring andmanagement of intracranial hypertension.Non-invasive methods for measuring ICP areunreliable in giving accurate measurements ofpressure. Levene et al reported that in a groupof 23 babies who had direct subarachnoid pres-sure recordings, knowledge of the actual ICPreading might have altered management to thebenefit of the patient in only 9%/o of cases.8 Theroutine use of ICP monitoring cannot berecommended.

ANTICONVULSANT TREATMENTConvulsions occur commonly after birthasphyxia and anticonvulsant drugs are themost widely used pharmacological agents in

613

on 17 April 2019 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.68.5_S

pec_No.612 on 1 M

ay 1993. Dow

nloaded from

614

the management of infants with birth asphyxia.The immature brain appears to develop con-vulsions at a significantly lower threshold thanthe more mature brain.12 Seizures cause adoubling in cortical metabolic rate and it hasbeen suggested that this causes furtherneuronal injury as a result of relative substratedepletion. In immature animal models, theinduction of status epilepticus caused amarked reduction in eventual brain weight andreduced cerebral DNA concentration.13

Despite these studies, controversy exists asto whether neonatal seizures simply reflectneuronal compromise or whether the convul-sions contribute to further neuronal loss.There are no human data to support the latterconcept. Nevertheless, the use of multiple anti-convulsants is very common in the asphyxiatedneonate and it is clear that exposure ofthe neo-natal brain to these drugs may be associatedwith adverse effects.14

There have been few controlled studies onthe use of anticonvulsants in the neonatalperiod. Goldberg et al treated a group ofasphyxiated infants with a continuous infusionof thiopentone and compared the outcomewith a similar group treated by standard anti-convulsants. 15 There were no significant differ-ences in outcome, but 14 of the 17 infantsgiven thiopentone treatment required inotropesupport for hypertension compared with onlyseven of 15 controls.

Although there is no scientific evidence thatanticonvulsants improve outcome after neo-natal convulsions due to birth asphyxia, it isclinically difficult not to treat infants with neo-natal convulsions. In my view, it is not neces-sary to abolish all convulsions, but treatmentshould be instituted for frequent (>3 convul-sions per hour) or prolonged convulsions (anyfit lasting ¢ 3 minutes). I recommend pheno-barbitone as the first line anticonvulsant (20mg/kg loading dose followed by 3 mg/kg every12 hours). If frequent or prolonged convul-sions continue then a second half loading dosecan be given (10 mg/kg). Clonazepam is usedas a second line anticonvulsant (100 ,ug/kgloading dose which can be followed by a con-tinuous infusion of 10 pug/kg/hour if necessary).Anticonvulsants can be stopped once theinfant is thought to be neurologically normalon clinical examination.There are no data on the clinical significance

of electroconvulsive seizure activity in asphyxi-ated infants with reference to subsequent out-come. Electroencephalographic monitoring ofinfants may give very important prognosticinformation (see below), but I do not believethat it is yet a useful clinical technique fordeciding which infants to treat or when to startanticonvulsant treatment.

New treatmentsIt is of considerable interest that the neuron isremarkably resistant to asphyxial insult andmay recover to generate electrical potentialseven after an hour of severe hypoxic-ischaemicinsult. Evidence from magnetic resonancespectroscopy studies in asphyxiated neonates

suggest that there is progressive and perma-nent degradation of high energy ATP mole-cules within the brain, but no abnormalitiescan be detected on magnetic resonancespectroscopy for up to 24 hours after theinsult.16 Measurement of cerebral blood flowvelocity by Doppler ultrasound has shown thatan abnormal signal which accurately predictssevere cerebral injury only becomes abnormal24 hours after the asphyxial event.17 Theseobservations support the evidence from animalmodels that asphyxia sets up a cascade of intra-cellular events which causes the neuron to diesome hours after the acute insult. The processof reoxygenation and reperfusion afterasphyxia appears to be the spark that ignitesthe fuse to eventual neuronal death. Recentresearch has suggested a number of pathwaysby which this damage occurs and therapeuticstrategies are suggested by this work. Twopathways that offer promise of protecting theperinatal human brain are (i) that which pre-vents free radical damage and (ii) that whichcauses the antagonism of excitatory neuro-transmitters.

FREE RADICAL INJURYA free radical is a highly energetic substancewhich contains an uneven number of electronsin its outer ring. Two free radicals, the super-oxide (02-) and the hydroxyl ('OH), aregenerated by oxygen metabolism. Their halflives are very short, but they may under certaincircumstances generate chain reactions of freeradicals which cause damage to cellular mem-branes. Naturally occurring free radicalscavengers exist to limit the production ofthese substances, but these may be over-whelmed after asphyxial compromise.

During hypoxic-ischaemic insult, freeradicals are produced by the process ofdegradation of ATP. On reperfusion of thetissues xanthine oxidase metabolises molecularoxygen to produce oxygen free radicals. Themain source of xanthine oxidase in the brain isthe endothelial cell. Free radicals are alsoproduced in the cortex after asphyxia as theresult of arachidonic acid metabolism. After anasphyxial injury of the immature animal,indomethacin has been shown to inhibit theproduction of free radicals in the brain.'8Allopurinol is a free radical scavenger and alsoinhibits the enzyme xanthine oxidase. Palmeret al have shown that treatment with allopuri-nol before an asphyxial insult reduces bothbrain swelling and structural damage in theperinatal brain.'9

GLUTAMATE RELATED INJURYGlutamate is an excitatory neurotransmitterthat is particularly ubiquitous in the developingbrain. The glutamate neuroreceptor is stimu-lated by at least three ligands, of which N-methyl-D-aspartate (NMDA) opens a receptoroperated channel which allows calcium toenter the neuron. Asphyxia causes excessiverelease of glutamate from the presynapticvesicles and inhibits uptake of glutamate from

Levene

on 17 April 2019 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.68.5_S

pec_No.612 on 1 M

ay 1993. Dow

nloaded from

Management of the asphyxiated full term infant

the synaptic cleft. This causes hyperstimula-tion of the glutamate receptors resulting inearly and late damage to the cell. Abnormalstimulation of the non-NMDA receptorscauses entry of excess sodium and water intothe cell shortly after the asphyxial injury lead-ing to cytotoxic oedema. Neuronal deathoccurs later and appears to be more closelyrelated to excessive stimulation of the NMDAreceptors which causes accumulation of toxicconcentrations of calcium within the neuron.This leads to a cascade of biochemical events,including activation of intracellular proteasesand lipases with the secondary effect of genera-tion of oxygen free radicals which in turn causefurther damage to intracellular membranes.A group of substances antagonise the

NMDA channel and protect the perinatalbrain after asphyxia.20 One such substance,MK-801, when given after an NMDA insultresulted in 95% protection of the brain.21 Evenwhen given 120 minutes after the insult therewas some protection.22 Unfortunately, MK-801 and otherNMDA receptor antagonists arehighly toxic. Magnesium appears to be anaturally occurring antagonist which has areceptor site deep within the calcium channel.It has been suggested that increasing the extra-cellular neuronal magnesium concentrationmay protect the brain against hypoxic-ischaemic insults by an action similar toNMDA receptor antagonists. It has beenshown that treatment with magnesiumsulphate up to an hour after excessive exposureto a NMDA-like compound protected theanimal against neurological sequelae.23Magnesium sulphate is a substance that hasbeen widely used for over 60 years in perinatalpractice as a treatment of premature labourand severe pre-eclamptic toxaemia. It seems tobe well tolerated by the fetus and newborninfant, although transient hypotonia andlethargy are commonly seen for a few daysafterwards.There are no published data on the use of

NMDA antagonists in the human neonate, butthe role of magnesium sulphate deserves activeconsideration as a neuroprotective agent foruse after birth asphyxia.

The role ofbrain imagingThere are two indications for imaging the brainof asphyxiated newborn infants. The first is todiscover a treatable complication of theasphyxia; most notably a subdural collection.In my experience, subdural haemorrhagesevere enough to require surgical treatmentoccurs in <5% of full term asphyxiated infants.An early ultrasound scan within 12 hours ofbirth will detect a midline shift if there is asignificant lesion. As there appears to be norational basis for the systematic managementof cerebral oedema, frequent ultrasound orcomputed tomography to detect this are of novalue.

Imaging the brain may be ofvalue in the pre-diction of outcome. Extensive areas ofdecreased radiodensity have been shown to bea good prognostic indicator of death or severe

handicap. Two studies have reported asensitivity of this finding of 90% and 91%respectively.24 25 It is of interest that prognosisusing computed tomography is reliable only inthe second week of life.26

PrognosisThe aggressive early management of theseverely asphyxiated infant may be temperedby consideration of the eventual outcome ofthe baby and the risk of severe disability. Thequestion relating to prognosis may arise at twodifferent times during the course of the infant'smanagement.

(1) WHEN SHOULD RESUSCITATION STOP?If an infant has no cardiac output after 10minutes of effective resuscitation then treat-ment should be abandoned. There is no con-sensus view as to how long resuscitation shouldcontinue if the baby has a good cardiac outputyet has failed to breathe spontaneously. Thereare reports in the literature of babies not estab-lishing spontaneous respiration by 20 minutesand surviving to be normal. In a Swedish study25% of surviving babies who had not breathedspontaneously by 20 minutes were withoutsignificant handicap.27The literature and individual experience

suggests that the prognosis for normal or nearnormal survival if the baby has not breathed for30 minutes after birth is poor. Peliowski andFiner have reviewed the literature and reportthe outcome of only 35 full term babies whodid not breathe spontaneously by 30 minutesand 24 (80%) died or were significantly handi-capped.28 Unfortunately, many of the data ontime to respiration are anecdotal and policiesbased on time to achieve spontaneous respira-tion must be recognised as being prone toserious prejudice.

Depression of Apgar scores are also arelatively poor predictor of adverse outcome.An overview of three studies reporting mortal-ity and morbidity in infants with Apgar scoresof 3 or less at five minutes showed that thiscarried an overall risk of mortality of 16%, butonly a 3% risk of handicap in survivinginfants.28 Another study reported that the bestpredictor of death or handicap was an Apgarscore of 5 or less at 10 minutes.29As it is usually inexperienced medical staff

who are called to the delivery suite to resus-citate asphyxiated babies, I advise that if thereis any doubt as to whether to continue resusci-tation the baby should be given the benefit ofthe doubt and transferred to a neonatal inten-sive care unit for further management andobservation. It may be possible to give a moreaccurate prediction of outcome later thanimmediately after birth (see below).

(2) WHEN SHOULD MECHANICAL VENTILATIONBE WITHDRAWN?Signs of irreversible cerebral injury may bedelayed for many hours and accurate andhonest prognostication may have to be delayed

615

on 17 April 2019 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.68.5_S

pec_No.612 on 1 M

ay 1993. Dow

nloaded from

Levene

up to 24 hours. The severity of HIE has beenshown to be the most accurate clinical predic-tor of outcome after birth asphyxia in full terminfants.29 Unfortunately, the maximal degreeofHIE may not be determined until the baby isa few days old and it therefore cannot be usedas an early predictor of outcome. A number oftests have been shown accurately to predictoutcome when performed within 24 hours oflife.

Experiments in the developing brain haveshown that asphyxia causes a sequence ofabnormalities in the electroencephalogram.This includes an initial period of depressedelectrical activity lasting approximately eighthours, followed by a period of epileptiformactivity and culminating in loss of intensity atall frequencies by 72 hours.30 The final patternof low activity correlated with severe neuronaldamage.31 In human neonates, similar severeabnormalities on the electrocephalogram orcerebral function monitor including burst sup-pression, sustained low voltage, and isoelectricactivity have also been shown to correlate withan overall risk of 95% for adverse outcome.28The time scale for these severe abnormalitiesto develop appears to be shorter in the humanthan in the animal model.

Severely abnormal evoked potentials alsoaccurately predict adverse outcome, althoughprediction of outcome in the first 24 hoursusing these techniques is not reliable.32 33 Wehave shown that abnormally high cerebralblood flow velocity detected by duplexDoppler ultrasound has a positive predictivevalue of 94% for adverse outcome (death orsevere handicap) when performed in the first24 hours of life. 17The presence of severely abnormal results

on electroencephalography or Doppler assess-ments in the first 24 hours appears to predictaccurately bad outcome and I advise that thesetests be repeated twice in the first 24 hours andif on both occasions the results are unequivo-cally abnormal then the poor prognosis shouldbe explained to the parents and considerationgiven to withdrawing ventilatory support if thisis appropriate.

1 Hull J, Dodd KL. Falling incidence of hypoxic-ischaemicencephalopathy in term infants. Br J Obstet Gynaecol1992; 99: 386-91.

2 Myers RE, Yamaguchi S. Nervous system effects of cardiacarrest in monkeys. Preservation of vision. Arch Neurol1977; 34: 65-74.

3 Vannucci RC, Mujsce DJ. Effect of glucose on perinatalhypoxic-ischemic brain damage. Biol Neonate 1992; 62:215-24.

4 Hattori H, Wasterlain CG. Posthypoxic glucose supplementreduces hypoxic-ischemic brain damage in the neonatalrat. Ann Neurol 1990; 28: 122-8.

5 Sheldon RA, Partridge JC, Ferriero DM. Postischemichyperglycemia is not protective to the neonatal rat brain.PediatrRes 1992; 32: 489-930.

6 Altman DI, Young RS, Yagel SK. Effects of dexamethasonein hypoxic ischemic brain injury in the neonatal rat. BiolNeonate 1984; 46: 149-56.

7 De Souza SW, Dobbing J. Cerebral oedema in developingbrain. III. Brain water and electrolytes in immature

asphyxiated rats treated with dexamethasone. Biol Neonate1973; 22: 388-97.

8 Levene MI, Evans DH, Forde A, Archer LNJ. Value ofintracranial pressure monitoring of asphyxiated newborninfants. Dev Med Child Neurol 1987; 29: 311-9.

9 Mujsce DJ, Stern DR, Vannucci RC, Towfighi J, HersheyPA. Mannitol therapy in perinatal hypoxic-ischemic braindamage. Ann Neurol 1988; 24: 338.

10 Marchal C, Costagliola P, Leveau P, Dulcq P, Steckler R,Rouquier F. Traitement de la souffrance cerebrale neo-natale d'origine anoxique par le mannitol. Revue Pediatrie1974; 9: 581-90.

11 Levene MI, Evans DH. Medical management of raisedintracranial pressure after severe birth asphyxia. Arch DisChild 1985; 60: 12-6.

12 Jensen FE, Applegate CD, Holtzman D, Belin TR,Burchfiel JL. Epileptogenic effect of hypoxia in theimmature rodent brain. Ann Neurol 1991; 29: 629-37.

13 Wasterlain CG. Effects of neonatal status epilepticus on ratbrain development. Neurology 1976; 26: 975-86.

14 Levene MI. Neonatal seizures. Neonatal neurology. Currentreviews in paediatrics. Edinburgh: Churchill Livingstone,1987: 201-38.

15 Goldberg R, Moscoso P, Bauer C, et al. Use of barbituratetherapy in severe perinatal asphyxia: a randomized con-trolled trial. JPediatr 1986; 109: 851-6.

16 Wyatt JS, Edwards AD, Azzopardi D, Reynolds EOR.Magnetic resonance and near infrared spectroscopy forinvestigation of perinatal hypoxic-ischaemic brain injury.Arch Dis Child 1989; 64: 953-63.

17 Levene MI, Fenton AC,. Evans DH, Archer LNJ,Shortland DB, Gibson NA. Severe birth asphyxia andabnormal cerebral blood-flow velocity. Dev Med ChildNeurol 1989; 31: 427-34.

18 Pourcyrous M, Leffler CW, Mirro R, Busija DW. Brainsuperoxide anion generation during asphyxia and reventi-lation in newborn pigs. Pediatr Res 1990; 28: 618-21.

19 Palmer C, Vannucci RC, Towfighi J. Reduction of perinatalhypoxic-ischemic brain damage with allopurinol. PediatrRes 1990; 27: 332-6.

20 Levene MI. Role of excitatory amino acid antagonists in themanagement of birth asphyxia. Biol Neonate 1992; 62:248-51.

21 McDonald JW, Roeser NF, Silverstein FS, Johnston MV.Quantitative assessment of neuroprotection againstNMDA-induced brain injury. Exp Neurol 1989; 106:289-96.

22 McDonald JW, Silverstein FS, Cardona D, Hudson C,Chen R, Johnston MV. Systemic administration of MK-801 protects against N-methyl-D-aspartate andquisqualate-mediated neurotoxicity in perinatal rats.Neuroscience 1990; 36: 589-99.

23 Wolf G, Keilhoff G, Fischer S, Hass P. Subcutaneouslyapplied magnesium protects reliably against quinolinate-induced N-methyl-D-aspartate (NMDA)-mediated neuro-degeneration and convulsions in rats: are theretherapeutical implications? Neurosci Lett 1990; 117:207-11.

24 Adsett DB, Fitz CR, Hill A. Hypoxic-ischemic cerebralinjury in the term newborn: correlation of CT findingswith neurological outcome. Dev Med Child Neurol 1985;27: 155-60.

25 Lipper EG, Voorhies TM, Ross G, Vannucci RC, Auld P.Early predictors at one-year outcome for infants asphyxi-ated at birth. Dev Med Child Neurol 1986; 28: 303-9.

26 Lipp-Zwahlen AE, Deonna T, Chrzanowski R, Micheli JL,Calame A. Temporal evolution of hypoxic-ischaemicbrain lesions in asphyxiated full-term newboms assessedby computerized tomography. Neuroradiology 1985; 27:138-44.

27 Ergander U, Eriksson M, Zetterstrom R. Severe neonatalasphyxia: incidence and prediction of outcome in theStockholm area. Acta Paediatr Scand 1983; 72: 321-5.

28 Peliowski A, Finer NN. Birth asphyxia in the term infant.In: Sinclair JC, Bracken MB, eds. Effective care of the new-born infant. Oxford: Oxford University Press, 1992:249-80.

29 Levene MI, Sands C, Grindulis H, Moore JR. Comparisonof two methods of predicting outcome in perinatalasphyxia. Lancet 1986; i: 67-8.

30 Williams CE, Gunn AJ, Mallard C, Gluckman PD.Outcome after ischemia in the developing sheep brain: anelectroencephalographic and histological study. Neurology1992; 31: 14-21.

31 Williams CE, Gunn AJ, Synek B, Gluckman PD. Delayedseizures occurring with hypoxic-ischemic encephalopathyin the fetal sheep. Pediatr Res 1990; 27: 561-5.

32 Gibson NA, Brezinova V, Levene MI. Somatosensoryevoked potentials in the term newborn. ElectroencephalogrClin Neurophysiol 1992; 84: 26-31.

33 Taylor MJ, Murphy WJ, Whyte HE. Prognostic reliability ofsomatosensory and visual evoked potentials of asphyxi-ated term infants. Dev Med Child Neurol 1992; 34:507-15.

616

on 17 April 2019 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.68.5_S

pec_No.612 on 1 M

ay 1993. Dow

nloaded from