Utility of Serial EEGs in Neonates During Extracorporeal Membrane Oxygenation L e o p o l d J. S t r e l e t z , M D * , M a r k D . B e j , M D * , L e o n a r d H e m a n t J. D e s a i , M D t , S a b r i n a G . B e a c h a m * ,
J. Graziani, MD*t,
James Cullen, RNt, and
Alan R. Spitzer, MDt
We found electroencephalographic (EEG) studies to be useful for monitoring cerebral function, for confirming seizure activity, and for limited prediction of shortterm outcome in 145 neonates who required extra-corporeal membrane oxygenation (ECMO) of reversible respiratory failure. The EEG tracings were classified as normal or as mildly, moderately, or markedly abnormal; abnormal recordings were further classified as focal, diffuse, or predominantly lateralized. A significant decrease in frequency and degree of EEG abnormalities was observed in recordings obtained after ECMO compared to those obtained prior to (P = .001) or during ECMO (P = .001). There was no significant increase in marked EEG abnormalities when recordings obtained before and during ECMO were compared (P = 0.41). Of 11 infants with electrographic seizures during ECMO, 7 (64%) either died during their nursery courses or were developmentally handicapped at age 1 year which is a significantly greater adverse outcome than that observed in infants without EEG seizure activity (P < .003). No consistently lateralized EEG abnormalities were observed during or after ECMO when compared to tracings obtained before cannulation of the right common carotid artery. There was no acute change in EEG rhythm or amplitude over the right cerebral hemisphere during right common carotid artery cannulation. Our observations support the value of serial EEG in the assessment of cerebral function in critically ill infants undergoing ECMO. They further suggest that, in this patient population, cannulation of the right common carotid artery is a safe procedure that does not result in lateralized abnormalities of cerebral electrical activity. Streletz L J, Bej MD, Graziani L J, Desai HJ, Beacham SG, Cullen J, Spitzer AR. Utility of serial EEGs in neonates during extracorporeal membrane oxygenation. Pediatr Neurol 1992;8:190-6.
From the Departments of *Neurology and *Pediatrics; Jefferson Medical College; Thomas Jefferson University; Philadelphia, Pennsylvania.
190 PEDIATRIC NEUROLOGY
Vol. 8 No. 3
Introduction Extracorporeal membrane oxygenation (ECMO) has been used since 1971 in the treatment of reversible respiratory failure in neonates [1-3]. During venoarterial ECMO, the right internal jugular vein is cannulated and venous blood from the right atrium is oxygenated externally. This blood is then returned to the aortic arch via the cannulated fight common carotid artery (RCCA). The vessels are permanently ligated in most infants at the time of cannulation , but reconstruction is possible . Usually, 60-80% of the cardiac output must flow through the bypass circuit to allow for adequate oxygenation. As the infant's pulmonary function improves, the ECMO flow is gradually decreased and is usually discontinued within 3 weeks. Survival rates of 72-92% have been reported in several large series of infants [6,71 who otherwise would have had a 20% chance of survival [8-10]. Infants treated with ECMO are initially pharmacologically paralyzed and are anticoagulated throughout the bypass procedure. They are at risk for hypoxia, cerebral ischemia, intracranial hemorrhage, and seizures. The incidence of post-ECMO epilepsy is believed to be low, but the incidence of seizures during ECMO is unknown . In some studies of ECMO-treated neonates, a predilection for right hemispheric structural or functional abnormalities was reported [10,12-18], but consistent laterality has not been reported by others [6,17-23]. We report our experience with serial EEG studies in the intensive care nursery (ICN) of Thomas Jefferson University Hospital (TJUH) in neonates undergoing ECMO. The purpose of these studies was to determine the following: the incidence of electrographic seizures and other EEG abnormalities in this high-risk population, the immediate effects of the ECMO surgical procedure (RCCA cannulation) on EEG activity, the prognostic significance of EEG background activity and lateralizing features, and the relationship of EEG abnormalities to clinical features of these infants.
Communications should be addressed to: Dr. Streletz; Electrodiagnostic Laboratory; Thomas Jefferson University Hospital; 11! South l lth Street; Suite G-9350; Philadelphia, PA 19107. Received November 7, 1991; accepted February 21, 1992.
Table 1. Primary diagnosis and mortality in 145 infants treated with ECMO Primary Diagnosis
Congenital diaphragmatic hernia
Hyaline membrane disease
Persistent pulmonary hypertension
Methods Patients. Between July, 1985 and December, 1990, 145 infants admitted to the ICN at TJUH met criteria for and were treated with ECMO  due to acute, severe respiratory failure. All infants were born at 36 weeks gestation or more. Informed parental consent was obtained for each infant. In this group, ECMO was initiated within the first week of life and was continued until respiratory function improved or until complications precluded continuation of the bypass. Only one infant died shortly after the initiation of the bypass procedure and before an EEG recording was obtained. Following discharge, all but 12 survivors were clinically evaluated. In this study, 145 term infants, 101 males (69.7%) and 44 females (30.3%) were evaluated. The mean gestational age was 39.1 _+ 2.2 (S.D.) weeks and mean birth weight 3.23 _+0.54 kg. All presented with acute respiratory failure and their primary diagnoses are listed in Table 1. ECMO was initiated within the first week of life and was continued for 2-14 days (151 + 79 hrs). Although all infants survived ECMO, 18 infants (12%) died immediately or shortly after termination of this procedure. EEG recordings were obtained prior to the bypass procedure in 67 infants, including 14 infants monitored during RCCA ligation and cannulation. EEGs were performed at least once in 139 infants during ECMO and 135 infants after ECMO. A complete series (pre-, during, and post-ECMO) of EEGs was obtained in 60 infants. Neonatal EEG Recordings. Scalp EEG recordings were obtained between 24 and 72 hours after ECMO was initiated and within 5 days after the completion of ECMO. EEGs were also performed prior to the initiation of ECMO, including monitoring of the RCCA ligation in 14 infants. The International 10/20 System modified for newborns was used and electrodes were attached with collodion. This technique has been described in detail elsewhere [24,25]. EEG criteria of Tharp and Laboyrie were used to grade each record as mild, moderate, or marked . These criteria were modified to take laterality of the abnormalities into consideration. Marked abnormalities were further tabulated as follows: isoelectric, paroxysmal (burst-suppression), generalized suppression, and generalized slowing, each classifted as absent or present. Electrographic seizure, focal suppression, focal slowing, and positive sharp waves were classified as left-predominant, right-predominant, or bilateral. Mild and moderate abnormalities, including other sharp transients, were similarly classified. All EEGs were interpreted and classified by one of the authors (LJS). Sharp EEG transients consisting of individual focal sharp waves, spikes, and repeated sharp discharges (< 300 ms) were considered to be a moderate EEG abnormality when present in excess of 2/min during quiet (non-REM) sleep . Electrographic seizure activity, usually beginning as low-amplitude activity in the 8-15 Hz range, slowing to
rhythmic, monomorphic waveforms at 0.5-3.0 Hz, and then stopping abruptly, was classified as a marked abnormality. Neuroimaging. Cranial ultrasonography (US) studies were obtained prior to ECMO, daily during ECMO, and then weekly during the nursery course of all infants, using methods previously reported . Unenhanced cranial computed tomographic (CT) scans were obtained prior to discharge in most survivors. Follow-up neurologic and developmental examinations were classified by clinical outcome, including developmental delay, cerebral palsy, epilepsy, and neonatal death. Developmental Assessment. All but 12 surviving infants were followed after discharge at regular intervals in an outpatient program at TJUH which consisted of outpatient visits 3, 6, 9, 12, 18, 24, 30, and 36 months after discharge and yearly thereafter. Children, ages 12-30 months, were evaluated with the Bayley Scales of Infant Development (BSID) . Those older than 30 months were evaluated with the Mullen Scales of Early Learning (MSEL) . The presence or absence of neurologic abnormalities, including asymmetry of motor function or reflexes, was observed. Cerebral palsy was defined as markedly delayed motor development associated with abnormal postures, muscle tone, and pathologic reflexes . Infants and young children who had nonspecific developmental delay without pathologic reflexes, abnormal postures, or muscle tone, or who eventually ambulated without spasticity, were not considered to have cerebral palsy. Standard scores on the mental and psychomotor indices of the BSID or MSEL were obtained at least once in all but 12 survivors between the ages of 6 months and 4 years. The examiners were unaware of the neonatal clinical, EEG, or neuroimaging findings. Standard scores were corrected for gestational age until age 12 months. Normal infants were those with all developmental scores within 2 S.D. of the mean. Developmental handicap was defined as any scale of the BSID or MSEL being < 2 S.D. below the mean, the presence of spastic cerebral palsy, or both. Data Analysis and Statistics. The gestational ages, birth weights, survival, and primary diagnoses were quantified as mean values with S.D.s and percentages. Severity and laterality of EEG abnormalities were compared by the ~2 statistic. Comparisons of the evolution of the EEG abnormalities before, during, and after ECMO were performed with the Wilcoxon Signed-Rank Test. Neurodevelopmental outcomes (normal versus developmentally handicapped) in infant groups were compared with the Fisher exact test. All results were computed using commercially available software. Because most of the infants were younger than 2 years of age at the last follow-up evaluation, an analysis of the relationship between EEG abnormalities and the Bayley Scale scores, other than the developmental handicap classification described above, was not attempted.
Serial EEGs and characterization of abnormalities. The classification of 338 EEGs (in 145 neonates) by the most severe abnormality is summarized in Table 2. Before and during ECMO therapy, the percentage of total EEG abnormalities (86%) did not differ. A slightly greater percentage of marked abnormalities was found during ECMO (47%) than before (42%), but did not reach statistical significance (P = 0.41). The EEG abnormalities before, during, and after ECMO consisted predominantly of moderate to marked abnormalities of the background activity which were bilateral and diffuse. The paroxysmal pattern was the most common of the marked EEG abnormalities and was observed both before and during ECMO (Table 3). Slow rhythms and suppression of the EEG background were observed mostly during ECMO. An isoelectric pattern was recorded in one infant during cardiopulmonary arrest shortly before RCCA cannulation. The ab-
Streletz et al: EEGs in ECMO-treated Infants
Table 2. Classification of EEGs during the neonatal course of ECMOtreated infants according to the most severe abnormality of each record EEG Classification Normal
Pre-ECMO (N = 67)
ECMO (N = 139)
Post-ECMO (N = 135)
normality persisted less than 24 hours, but the infant had significant neurologic morbidity at the follow-up examination. One other infant with an isoelectric EEG pattern did not survive the neonatal period. EEG abnormalities were either persistent or transient; the latter were generally regarded as epileptiform abnormalities and are described below. Persistent EEG abnormalities consisted of focal slowing or lateralized suppression, which had to be distinguished from factitious asymmetries by serial recordings. Such background asymmetries were often caused by scalp edema resulting from the dependent position of the head during ECMO. Laterality of EEG abnormalities. Although most EEG abnormalities were bilateral or generalized, many abnormalities were focal or predominantly lateralized. Most infants with iateralized EEG abnormalities had no neuro-
imaging abnormalities, but structural abnormalities on cranial US and CT were occasionally observed (Fig 1). In infants with marked EEG abnormalities, a slight preponderance of right- versus left-sided EEG abnormalities was found overall (36 vs. 23, P > .05). These findings, however, were outweighed by bilateral abnormalities (Table 3). Serial recordings in individual infants failed to demonstrate persistence of the lateralized features. Moderate to marked background slowing, unilaterally and bilaterally, was more frequent before and during than after ECMO. No lateralized changes in EEG patterns or rhythms were observed during RCCA ligation in the 14 infants monitored during cannulation.
Sharp Transients, Seizures, and Other Epileptiform Abnormalities. Sharp transients were most abundant in the centrotemporal leads and often had right-sided predomi-
Table 3. Marked EEG abnormalities during the neonatal course of 86 infants who required ECMO EEG Classification
2 0, I, 1
Positive Sharp Waves:
* The number of marked EEG abnormalities exceeds the total number of infants with marked EEG records (Table 2) because some EEGs contained more than one marked abnormality. Abbreviations: B = Bilateral, without hemispheric predominance L = Left hemispheric predominance R = Right hemispheric predominance
Vol. 8 No. 3
L EOM ~
R EOM EMG
F P l C3
C3 01 FP2 C 4 - -
C40 2 ~
A1 Ta , ' ~ - - ' ~ W " ~ / ' ~ ' T3 C 3
%Cz C z C4
- I,,,, _ .
Figure 1. EEG with right-sided suppression of the background following cerebral infarction. Subsequent EEG revealed evolution to hemi-hypsarrhythmia pattern with infantile spasms.
nance before and after ECMO (Fig 2). They were usually not associated with a structural lesion or neurologic deficit unless accompanied by a more severe EEG abnormality. The electrographic seizures had no consistently lateralizing features but were often multifocal or occasionally midline (Table 3). Eighteen EEG seizures were observed in 16 ECMO-treated neonates (9%) during their neonatal courses, but convulsive seizures were rarely observed (Fig 3). The lack of clinical signs of seizure activity was usually due to the use of neuromuscular blocking agents during the period of most severe respiratory distress. j
Positive sharp waves were infrequently observed . When present, they were usually accompanied by other, more severe EEG abnormalities and by US structural lesions. As was observed with electrographic seizures, positive sharp waves were most often observed during ECMO (Table 3). EEG Studies and Prognosis. EEG tracings obtained before, during, and after ECMO were compared in 60 infants who had completed studies (Table 4). A statistically significant improvement in EEGs after ECMO was observed compared to recordings obtained prior to the bypass ~'
•, ¢3- CZ . O I - T4
~ Im.V 2me.
Figure 2. EEG reveals interictal spike-and-sharp-wave discharges (> 2/min) bilaterally and independently in the central and temporal regions of both cerebral hemispheres. No clinical seizures were observed.
Streletz et ah EEGs
. . . . . . . .
. . . . .
= . . = = . =
. . . .
. . . . = = . ° . o .
. . . .
. ~ . . . . = . . . .
, . . . . . . . . . . . . . . . . . . . . . . . .
. . . .
= . = . . = ~ |
. . . .
REOM ~ ~ . . . ~ . , . . ~ : . . ~ EMG
. . . . .
. . . . - . , , . d . . , _ ~ : , , . ~ . . , : • - :. - ~ ~ . ~ , ~ . ~ . .
: ..... :4"~
i .... =..........................................
c3 bl -,_-
Fp= C4 C4
^~ Ts T3 C 3
%% c z c4 c 4 T4 T4A 2 Seizure
Figure 3. EEG seizures detected during ECMO in an infant with a paroxysmal background. Note midline Cz origin of seizure. The infant was paralyzed with a neuromuscular blocking agent and no behavioral correlate was observed.
(Wilcoxon Signed-Rank Test, P = .0001). Neurologic outcome was good in infants with normal or mildly abnormal EEGs before and during ECMO. Sixteen of 18 infants (88%) who died shortly after ECMO had moderately to severely abnormal EEGs recorded at least once during their nursery courses. Many infants with moderate and marked E E G abnormalities demonstrated resolution during ECMO, limiting their prognostic value. One EEG feature (i.e., EEG seizures) did prove prognostically useful. Table 4 compares the neurodevelopment of infants with (N = 11) and without (N = 114) electrographic seizures during ICN admission and lists those who either died or were examined at age 1 year or older. Of the 114 infants without electrographic seizures, 10 were lost to follow-up before the age of 1 year and 4 had genetic brain defects. The remaining 100 infants were included in the analysis summarized in Table 5. In a dichotomous comparison of normal infants to dead or handicapped inT a b l e 4. C o m p a r i s o n o f p r e - E C M O to p o s t - E C M O E E G g r a d e s of 60 serially studied surviving neonates ~
Post-ECMO EEG* Mild Moderate
fants (Table 5), EEG seizures during ECMO were significantly related to poor outcome (Fisher exact test, P < .003). Two of the survivors who had neonatal electrographic seizures have cerebral palsy and continue to have seizures following discharge and require antiepileptic medication.
Discussion Although ECMO reduces mortality in appropriately selected infants with respiratory failure, neurologic deficits of uncertain etiology have been observed in a significant number of survivors [ 15-17]. In a retrospective review utilizing CT scans, EEG, and follow-up neurologic examinations, Schumacher et al. reported that 8 of 59 ECMO survivors had right hemispheric brain abnormalities [ 15]. The authors proposed that carotid artery ligation was not without risk, particularly in infants with hypoxemia, but made no comparison with left-sided or bilateral abnormalities. Campbell et al. evaluated 35 infants treated with
T a b l e 5. Clinical o u t c o m e in i n f a n t s a g e s 1 y e a r o r older with and without electrographic seizures during ECMO* Clinical O u t c o m e
Vol. 8 No. 3
Developmental handicap or neonatal death
* Wilcoxon signed-rank test, P = .0001. * Improved: 36, unchanged: 16, deteriorated: 8.
* Fisher exact test, P < .003. ~ Two infants died. Sixteen infants died.
ECMO and observed a significantly higher incidence of left (9 of 35) versus right (2 of 35) focal clinical seizures ; yet, CT scans, US studies, EEG tracings, and neurologic examinations did not reveal an increased incidence of right hemispheric abnormalities. Andrews et al. reported right-sided weakness in a venovenous ECMO-treated infant and left-sided weakness in another infant whose right carotid artery was cannulated . Both infants, and one other who developed spastic quadriparesis, had significant hypoxia, ischemia, or both prior to ECMO therapy. Lott et al. reported no significant lateralized EEG abnormalities but did observe a reduction in the amplitude of right hemispheric long-latency components of the auditory and somatosensory evoked potentials in 10 survivors of ECMO . Utilizing nuclear MRI, Wiznitzer et al. reported brain parenchymal abnormalities in 33% of infants who survived ECMO, although there was no increased frequency of right hemispheric lesions . A possible mechanism related to right hemispheric vascular abnormalities observed in some ECMO infants is hypoxia combined with cannulation and ligation of the RCCA. In animal studies, Rice et al. reported that unilateral ligation of the common carotid artery in combination with hypoxia and low cardiac output reduced cerebral perfusion and resulted in ipsilateral brain atrophy . In contrast to reports of a possible predilection for right hemispheric vascular lesions in some ECMO patients, reports of neonatal cerebral infarcts in non-ECMO-treated infants have noted that about 70% occur in the left hemisphere [33-37]. There appears to be no predilection for consistently lateralized hemispheric damage in infants with persistent pulmonary hypertension managed without ECMO [38,39]. We observed that, by EEG criteria, neither right nor left hemispheric abnormalities predominated, but most were bilateral or generalized in distribution. Further, EEG tracings obtained shortly after ECMO were significantly improved compared with those recorded prior to or during ECMO therapy. In most infants, EEG abnormalities during ECMO were associated with normal neurosonograms during and following the procedure, suggesting that gross structural brain abnormalities are not necessarily the cause of the abnormal EEG recording. In addition, EEG abnormalities occurred prior to cannulation of the RCCA in the majority of infants studied before ECMO. Our observation of electrographic seizures during ECMO (9% of neonates) demonstrates the clinical utility of serial EEG studies in this population. Because the incidence of neonatal seizures that may require treatment is reported to be as high as 70% in asphyxiated neonates [26,40-42], EEG monitoring for the identification of epileptic activity is appropriate management of infants during ECMO. In a recent study, Jensen et al. reported a marked susceptibility of the hypoxic immature rodent brain to the development of epileptiform EEG activity and seizures ; however, clinical observation alone may not be entirely reliable in infants who are paralyzed with pancur-
onium during the acute phase of respiratory failure and ECMO therapy . We found that the identification of neonatal seizures was significantly correlated with a poor outcome in infants treated with ECMO, an observation consistent with the recent reports of Hofkosh et al.  and Campbell et al. . Although our neurodevelopmental follow-up is still incomplete, our results are consistent with previous reports [ 11,46] of a poor prognosis but a low incidence of postnatal epilepsy in infants who had EEG seizures associated with ECMO. The fact that most EEG abnormalities were nonlateralized and improved after ECMO suggests that the acute effects of respiratory failure are temporally related to brain electrical activity. EEG abnormalities during and following ECMO also probably reflect the adverse effects of pre-existing perinatal complications as evidenced by our findings that 80% of infants studied prior to ECMO had moderately to severely abnormal recordings before cannulation; however, in our study, we were unable to define further the prenatal and neonatal factors related to the possible pathogenesis of the EEG abnormalities observed prior to ECMO. Further study and follow-up will be required to determine their ultimate prognostic value. Our results to date reveal no deleterious effects either of permanent RCCA ligation or of the ECMO procedure p e r se in term neonates. On the contrary, we believe that our results are evidence that this procedure is of significant benefit to selected neonates with severe respiratory failure.
This research was supported by the National Institute of Health, Grant HD 21453 and NS 27463. The authors thank Gretchen McCann for editing, as well as Paul Micou and Joanna Turner for data filing and compilation. We also thank the medical and nursing staff of the Intensive Care Nursery, in particular Drs. Shobhana A. Desai, Nancy Robinson, and Michael Kornhauser, and the EEG technologists of the Electrodiagnostic Laboratory. Christian Stanley, RN and Dr. Gabriel Tatarian provided follow-up and statistical analysis.
 Bui KC. Extracorporeal membrane oxygenation in neonatal respiratory failure: A brief overview. J Perinatol 1989;9:323-6.  White JJ, Andrews HG, Risemberg H, Mazur D, Hailer JA Jr. Prolonged respiratory support in newborn infants with a membrane oxygenator. Surgery 1971;70:288-96.  Short BL, Lotze A. Extracorporeal membrane oxygenator therapy. Pediatr Ann 1988;17:516-23.  Bartlett RH, A n d r e w s AF, Toomasian JM, Haiduc NJ, Gazzaniga AB. Extracorporeal membrane oxygenation for newborn respiratory failure: Forty-five cases. Surgery 1982;92:425-32.  Cromblehohne TM, Adzick NS, deLorimier AA, Longaker MT, Harrison MR, Charlton VE. Carotid artery reconstruction following extracorporeal membrane oxygenation. Am J Dis Child 1990;144:872-4. [6l Glass P, Miller MK, Short BL. Morbidity for survivors of extracorporeal membrane oxygenation: Neurodevelopmental outcome at 1 year of age. Pediatrics 1989;83:72-8.  Bartlett RH, Toomasian J, Roloff DW, Gazzaniga AB, Corwin AG, Rucker R. Extracorporeal membrane oxygenation (ECMO) in neonatal respiratory failure: 100 cases. Arm Surg 1980;204:236-45.
Streletz et al: EEGs in ECMO-treated Infants
 Bartlett RH, Roloff DW, Cornell RG, Andrews AF, Dillin PW, Zwischenberger JB. Extracorporeal circulation in neonatal respiratory failure: A prospective randomized study. Pediatrics 1985;76:479-87.  Beck R, Anderson KD, Pearson GD, Cronin J, Miller MK, Short BL. Criteria for extracorporeal membrane oxygenation in a population of infants with persistent pulmonary hypertension of the newborn. J Pediatr Surg 1986;21:297-302.  K r u m m e l TM, Greenfield LJ, Kirkpatrick BV, et al. The early evaluation of survivors after extracorporeal membrane oxygenation for neonatal pulmonary failure. J Pediatr Surg 1984;19:585-90.  Horton EJ, Lew AD, Ramos AD, et al. Seizure, epilepsy, and ECMO graduates. Epilepsia 1990;31:612. [121 Voorhies TM, Tardo CL, Starrett AL, et al. Evaluation of the cerebral circulation in neonates following extracorporeal membrane oxygenation. Ann Neurol 1985;18:380.  Miller MK, Glass P, Anderson KD, Short BL. Neurodevelopment of moribund neonates treated with extraeorporeal membrane oxygenation (ECMO). Pediatr Res 1986;20:381A.  Mitchell DG, Merton DA, Graziani LJ, et al. Right carotid artery ligation in neonates: Classification of collateral flow with color Doppler imaging. Radiology 1990;I 75:117-23.  Schumacher RE, Barks JDE, Johnston MV, et al. Right-sided brain lesions in infants following extracorporeal membrane oxygenation. Pediatrics 1988;82:155-61.  Andrews AF, Nixon CA, Cilley RE, Roloff DW, Bartlett RH. One-to-three year outcome for 14 neonatal survivors of extracorporeal membrane oxygenation. Pediatrics 1986;78:692-8.  Campbell LR, Bunyapen C, Holmes GL, Howell CG, Kanto WP Jr. Right common carotid artery ligation in extracorporeal membrane oxygenation. J Pediatr 1988;113:110-3.  Lott IT, McPherson D, Towne B, Johnson D, Starr A. Longterm neurophysiologic outcome after neonatal extracorporeal melnbrahe oxygenation. J Pediatr 1990;116:343-9.  Grazlani LJ, Streletz LJ, Kornhauser MS, et al. Electroencephalographic and cranial ultrasound studies in neonates treated with extracorporeal membrane nxygenation. Pediatr Rev Commun 1989;4: 71-5.  Taylor GA, Fitz CR, Miller MK, Garin DB, Catena LM, Short BL. Intracranial abnormalities in infants treated with extracorporeal membrane oxygenation: Imaging with ultrasound and computed tomography. Radiology 1987; 165:675-8. I21] Slovis TL, Sell LL, Bedard MP, Klein MD. Ultrasonographic findings (CNS, thorax, abdomen) in infants undergoing extracorporeal oxygenation therapy. Pediatr Radiol 1988; 18:112-7.  Towne BH, Lott IT, Hicks DA, Healey T. Long-term follow-up of infants and children treated with extracorporeal membrane oxygena tion (ECMO): A preliminary report. J Pediatr Surg 1985;20:410-4.  Wiznltzer M, Masaryk TJ, Lewin H, Walsh M, Stork E. Parenchymal and vascular magnetic resonance imaging of the brain after extracorporeal membrane oxygenation. Am J Dis Child 1990;144: 1323-6.  Beacham SG, Streletz LJ. Electroencephalographic studies in neonates during extracorporeal membrane oxygenation. Am J EEG Technol 1991;31:11-25. I25] Streletz LJ, Graziani LJ. Electroencephalography and evoked potentials in the neonate. In: Nelson NM, ed. Current therapy in neonatal-perinatal medicine. Philadelphia: BC Decker, 1985;329-34.
Vol. 8 No. 3
 T h a r p BR, Laboyrie PM. The incidence of EEG abnormalities and outcome of infants paralyzed with neuromuscular blocking agents. Crit Care Med 1983;11:926-9.  Hughes JR, Fino J, Gagnon L. The use of the electroencephalogram in the confirmation of seizures in premature and neonatal infants, Neuropediatrics 1983; 14:213-9.  Pidcock FS, Graziani L J, Stanley C, Mitchell DG, Merton D. Neurosonographic features of periventricular echodensities associated with cerebral palsy in preterm infants. J Pediatr 1990; 116:417-22.  Bayley N. Bayley scales of infant development: Birth to two years. New York: Psychological Corporation, 1969.  Mullen EM. Mullen scales of early learning. Cranston: T.O.T.A.L. Child, 1989.  Scher MS. Midline electrographic abnormalities and cerebral lesions in the newborn brain. J Child Neurol 1988;3:135-46.  Rice JE lIl, Vannucci RC, Brierley JB. The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 1981; 9:131-41.  Mannino FL, Tranner DA. Stroke in neonates. J Pediatr 1983; 102:605-10. [341 Mantovani JF, Gerber GJ. "'Idiopathic" neonatal cerebral infarction. Am J Dis Child 1984;138:359-62.  Ment LR, Duncan C, Ehrenkranz RA. Perinatal cerebral infarction. Ann Neurol 1984;16:559-68.  Levy SR, Abroms IF. Marshall PC, Rosquette EE. Seizures and cerebral infarction in the full-term newborn. Ann Neurol 1985;17: 366-70. I37] Claney RR, Malin S, Laraque D. Baumgart S, Younkin D. Focal motor seizures heralding stroke in full-term neonates. Am J Dis Child 1985:139:601-6.  Klesh KW, Murphy TF, Scher MS, Buchanan DE, Maxwell ER Guthrie RD. Cerebral infarction in persistent pulmonary hypertension nf the newborn. Am J Dis Child 1987;141:852-7. [391 Seher MS, Klesh KW, Murphy TF, Guthrie RD. Seizures and infarction in neonates with persistent pubnonary hypertension. Pediatr Neural 1986:2:332-9.  Clancy RR, Legido A. Postnatal epilepsy after EEG-confirmed neonatal seizures. Epilepsia 1991 ;32:69-76. [411 T h a r p BR, Cukier F, Monod N. The prognostic value nf the electroencephalogram in premature infants. Electroencephalogr Clin Neurophysiol 1981;51:219-36. [421 Rowe JC, Holmes GL, Hafford J, et al. Prognostic value of the electroencephalogram in term and preterm infants following neonatal seizures. Electroencephalogr Clin Neurophysiol 1985;60:183-96.  Jensen FE, Applegate CD, Holtzman D, Belin TR, Burchfiel JL. Epileptogenic effect of hypoxia in the immature rodent brain. Ann Neurol 1991 ;29:629-37. 1441 HotKosb D, Thompson AE, Nozza RJ, Kemp SS, Bowen A, Feldman H. Ten years of extracorporeal membrane oxygenation: Neurodevelopmental outcome. Pediatrics 1991;87:549-55.  Campbell LR, Bunyapen C, Gangarosa ME, Cohen M, Kanto WR Significance of seizures associated with extracorporeal membrane oxygenation. J Pediatr 1991 ;I 19:789-92. [461 Conry JA, Miller MK, Glass R Short BL. Neonatal seizures and EEGs in infants following extracorporeal membrane oxygenation (ECMO). Pediatr Res 1987;21:490A.