January 1981

112 The Journal o f P E D I A T R I C S

Hypoxic-ischemic encephalopathy in term neonates:

Perinatal factors and outcome

Ninety-five infants of 3 7 weeks" gestation or greater with evidence of hypoxic-ischemic encephalopathy

Jbllowing perinatal asphyxia were prospectively identified in the neonatal period. The degree of
encephalopathy was graded using the staging system of Sarnat and Sarnat. Six infants died, 78 infants
were sequentially followed in the Neonatal Follow-up Clinic, and in five additional infants, follow-up
information was available. The mean duration of follow-up was 19.3 months. Fifty-eight (65%) of the
89 infants followed were normal or mildly handicapped, six (7%) died, and the remainder had significant
handicap. There was no stgnificant relationship between any o f over 100 obstetrical antepartum or
intrapartum variables and outcome. Infants with five-minute Apgar scores of 0 to 3, seizures within the
first day of life, Stage 11 or H I encephalopathy, or a suppressed electroencephalogram had a significantly
greater incidence of severe handicap or death. In addition, although there were fewer females, they had
a significantly greater incidence of handicap. There appeared to be an improved outcome in the last two
years (1977-1978) compared to the first two years (1975-1976), suggesting that improved recognition and
neonatal management may lead to a decrease in significant sequelae.

N. N. Finer, M . D . , F.R.C.P.(C),* C. M . R o b e r t s o n , M . D . , F.R.C.P.(C),

R. T. Riehards, M.D., F.R.C.S.(C), L. E. Pinnell, R.N., M.N., and

K. L. Peters, R.N., B.N.Sc., E d m o n t o n , A l t a . , C a n a d a

PERINATAL ASPHYXIA usually refers to an insult tion evidenced by low Apgar scores, which further com-
accompanied by decreased oxygen delivery to the fetal or promises the hypoxic-ischemic insult. Following delivery,
neonatal brain. When asphyxia is followed by abnormal these infants display alterations in consciousness, muscle
neonatal behavior, a syndrome has been described known tone, and primitive reflexes, producing a recognizable
as "hypoxic-ischemic encephalopathy. ''1,~ The hypoxic- syndrome.
ischemic insult may result from impaired placental gas
exchange or blood flow, from umbilical cord compression, See related article, p. 106.
or may occur postnatally as a result of neonatal respirato-
ry or cardiac compromise. Postnatal insults usually
Abbreviations used
account for only 10% of infants with evidence of HIE? HIE: hypoxic-ischemic encephalopathy
Significant intrapartum asphyxia usually results in the EEG: electroencephalogram
birth of an infant with depressed cardiorespiratory func- CT: computerized tomography

Amiel-Tison4 and Brann and Dykes ~'demonstrated that

From the Northern and Central Alberta Perinatat
Program; Departments o f Neonatalogy and
Perinatology, Royal ,4 lexandra Hospital," Neonatal
Follow-Up Clinic, Glenrose School Hospital; and
Departments of Pediatrics and Obstetrics, University
of ,4lberta Medical School.
*Reprint address: Department Of Neonatology, Royal
Alexandra Hospital, 10240 Kingsway Ave., Edmonton,
Alberta T5H 3V9.
constant seizures or persistently abnormal electroenceph-
alograms increased the likelihood of later sequelae in
term infants with neurobehavioral evidence of a previous
hypoxic-ischemic injury. Brown et al:' showed that the
occurrence of perinatal asphyxia with certain behavioral
and neurologic signs in the neonatal period, such as
feeding difficulties, apneic or cyanotic spells, or apathy,

VoL 98, No. 1, pp. 112-117 0022-3476/81/010112+06500.60/0 9 1981 The C. V. Mosby Co.
Volume 98
Number 1
was associated with a significant risk of subsequent
handicap. In that study, 48% of the infants died or were
severely handicapped, and a further 16% showed evidence
of a minimal cerebral disorder.
In an effort to examine further the relationships of both
perinatal and neonatal events to outcome, 95 term neo-
nates with evidence of HIE were prospectively identified
and followed for up to four years.
METHODS AND MATERIALS
The 95 infants studied were all term singleton infants of
37 weeks' gestation or greater and admitted to the Royal
Alexandra Hospital Newborn Intensive Care Unit from
September, 1974, through December, 1978. Sixty-seven
infants (70.5% of the study population) were delivered in
the Royal Alexandra Hospital during the study period
among a total of 20,155 live births. Of the 28 transported
infants, 11 (39%) were transported within 12 hours of
birth, seven (25%) between 12 to 24 hours of birth, and 10
(36%) were transported after the first 24 hours of life.
The criteria for inclusion in the study were an abnormal
neurologic examination during the first week of life, and
one or more of the following criteria: (1) The documenta-
tion of intrapartum fetal distress through the recognition
of abnormal heart rate patterns ~ (n -- 35) with or without
the passage of meconium. (2) The presence of immediate
neonatal distress as evidenced by a low (less than 5)
one-minute (n = 84) or five-minute Apgar score
(n = 58). (3) The need for immediate neonatal resuscita-
tion, including bag and mask ventilation ( n - - 2 5 ) or
intubation (n = 9), with or without prolonged ventilation
for greater than five minutes (n = 19).
An abnormal neurologic examination included the
presence of any two or more of the following: (1)
Alterations of consciousness assessed following vigorous
stimulation to attempt to arouse the infant. (2) Alterations
of muscle tone categorized as global hypotonia or hyper-
tonia, as well as localized tone disturbances. (3) Abnormal
primitive reflexes including the Moro, grasp, and suck
responses. This examination was performed within one
hour of admission of the infant and was repeated daily by
an attending neonatologist.
From.October, 1976, onward, the Sarnat classification~
was utilized for defining the clinical stages of encephalo-
pathy (without electroencephalographic criteria). Briefly,
this classification involves three stages, with Stage I
characterized by hyperalertness, hyper-reflexia, dilated
pupils, tachycardia, and the absence of seizures. Stage II
includes the presence of lethargy, hyper-reflexia, miosis,
bradycardia, seizures, hypotonia, and a weak suck and
Moro response. Stage III is characterized by stupor,
flaccidity, small to midposition pupils which react poorly
Hypoxic-ischemic encephalopathy 1 13
to light, decreased stretch reflexes, hypothermia, and
absent Moro and suck responses.
Infants studied prior to October, 1976, were retrospec-
tively scored using their neurologic evaluation so as to fit
the Sarnat profile.' The most severe stage seen during the
first week in the infant was the stage assigned to that
infant.
Seizures included observed single or repetitive tonic or
clonic movements, as well as subtle seizures as defined by
Volpe. ~
No infant with a confirmed central nervous system
malformation, chromosomal abnormality, familial illness,
or intrauterine infection was included in the study. All
infants fulfilling the above criteria during the study period
were included.
Electroencephalograms were analyzed by the same
individual (C. R.), who was unaware of the infant's
clinical status at the time, and the background pattern was
interpreted as normal or suppressed. A suppressed back-
ground included electrocerebral silence, burst suppression
(periods of inactivity below 5/~V interrupted by bursts of
activity) or a low voltage background pattern with ampli-
tudes of 5 to 15 ~tV.7
Information regarding outcome of all children was
obtained through the Neonatal Follow-Up Clinic at the
Glenrose School Hospital in Edmonton, to which all
surviving infants were referred for long-term follow-up.
The infants were scheduled to be seen at the following
ages: 6 months, 12 months, 27 months, 31/2years, 51/2years,
and 8 years, using the norms for developmental levels
adapted from Gesell,~ Illingworth," and Mecham. ~~Cogni-
tive evaluations were performed using the Uzgiris-Hunt
scales of infant psychologic development" during the first
two years. At three and one-half years a Stanford-Binet
Intelligence Scale': and the Berry Development Test of
Visual Motor Integration':' were performed.
All demographic, obstetric, and neonatal data were
collected retrospectively through detailed review of the
hospital medical record, utilizing a medical data form
constructed specifically for this purpose. (Available from
authors upon request.)
Five categories of outcome were defined for the study
population: (1) Normal-within the average range of
development for age and no evidence of handicap. (2)
Mild handicap-variations from normal on neurologic or
developmental examination without a specific diagnosis.
(3) Moderate handicap-trainable retardation, severe
behavioral disorders, convulsive disorders (excluding
febrile convulsions), mild or moderate neurosensory deaf-
ness, spastic diplegia, hemiplegia, or visual impairment.
(4) Severe handicap-spastic quadraplegia, severe psycho-
motor retardation, or severe neurosensory deafness or
1 14 Finer et al. The Journal of Pediatrics
January 1981

Table I. Apgar scores presentations such as face and shoulder. Delivery was
m I spontaneous in 30 (32%) with forceps applied in 28 (29%),
Apgar score I One minute I Five minutes
of which 12 were midforceps applications for fetal dis-
I I
0-3 44 (47.3%) 14 (14.9%) tress. There were 33 cesarean sections (35%); one was
4-7 40 (43.0%) 44 (46.8%) elective and the remainder were emergent for fetal or
8-10 9 (9.7%) 36 (38.3%)
maternal distress.
Neonatal. There were 56 male infants and 39 females.
All infants were 37 weeks of age or greater by maternal
Table II. Neonatal complications
dates and/or Dubowitz examination.TM The infants' mean
Complication [ Incidence [ % birth weight was 3,080 gm (range 2,010 to 4,440 gm) and
66 were considered appropriate for gestational age, one
Convulsions during the first 7 days 65 68.4
infant was large for gestational age. Intrauterine growth
Age of onset
1 hr or less 19 20.0 retardation was diagnosed in 28 by physical examination,
2-24 hr 29 30.5 maternal dates, and gestational assessment using the
25-96 hr 13 13.7 examination of Dubowitz et al. TM The Apgar scores at one
97 hr-7 days 4 4.2 and five minutes are shown in Table I; the mean
Meconium aspiration 46 48.4
one-minute Apgar score was 3.8 and the mean five-
Mechanical ventilation 33 34.7
Oliguria 24 25.3 minute score was 6.3. Neonatal complications are listed in
Intracranial hemorrhage (proven or sus- Table II, in order of their frequency.
pected) 18 18.9 Electroencephalograms were performed in 51 infants
Subdural 3 3.3 within the first five days of life; 31% were normal. Of the
Subarachnoid 11 11.6
remainder, 16% were interpreted as showing a low voltage
Intraventricular 1 1.1
Unspecified 3 3.2 background with or without spike waves, 37% revealed
Hyperbilirubinemia (> 12 mg/dl) 17 17.9 burst suppression, and 16% had a normal background
with sharp waves. Thirty-seven infants had subsequent
EEGs at 14 to 30 days of life, of which 13 were normal.
blindness. (5) Death-death during hospitalization or Sixty-five infants received drug therapy specifically for
during the follow-up period. seizures or for evidence of increased intracranial pressure
Statistical evaluation consisted of the use of chi-square during the neonatal period. Phenobarbital was given to
evaluation for all obstetrical and neonatal variables in 62% of the infants, either alone (20%) or with other
relation to asphyxia staging and outcome. The level of medications (42%). Diphenylhydantoin and dexametha-
significance chosen was a P value of less than 0.05. sone were the next most frequently administered medica-
tions (32% and 24%), and were used only in combination
PATIENT POPULATION with phenobartial, and occasionally in addition to other
Obstetrical. The mean maternal age was 25 years agents (diazepam, mannitol, paraldehyde).
(range 16 to 38 years). There were 52 (55%) primigravid The staging of the HIE revealed that 35% were Stage I,
mothers. The average weight gain during pregnancy was 50% were Stage II, and 15% were Stage Ili.
11.4 kg (range 0 to 24 kg). Six of the mothers received no Outcome data were available for 89 of the 95 children
antenatal care. Antepartum hemorrhage (26%), urinary studied (94%). Six of these d i e d - o n e of sudden infant
tract infection (13%), hypertension (8%), pre-eclampsia death and one from aspiration secondary to neurologic
(6%), and eclampsia (7%) were the most common compli- handicap; the remaining four died as a direct result of
cations observed, severe neurologic damage in the early or late neonatal
Continuous intrapartum fetal monitoring-internal, period. Seventy-eight children were seen in the follow-up
external, or both-was carried out on 37 (39%) patients. clinic, and outcome information was available on five
Only two were completely normal; the remainder showed additional children through correspondence with the
late or late variable decelerations (54%), decreased vari- primary physician. The mean duration of follow-up was
ability (19%), baseline tachycardia (16%), or bradycardia 19.3 months with a range of 3 months to 4.3 years.
(24%).6 The mean length of labor was 7.7 hours (range 0 to
32 hours); 21 labors lasted from 9 to 24 hours and seven RESULTS
more than 24 hours. The overall results are shown in Table III. Sixty-five
Seventy-nine (83%) of presentations were occipital, 13 percent of the infants followed were normal or minimally
(14%) were breech, and three (3%) represented other handicapped, 28% were significantly handicapped, and
Volume 98
Number 1
Table III. Outcome by A p g a r scores at five minutes, age at onset o f convulsions, and
encephalopathy stage
Mild
Normal handicap
0-3 4 I
4-10 39 13
Convulsions
<24 hr 17 7 9
24-168 hr 12 1 1
Encephalopathy stage
Mild (1) 19 6 0
Moderate (2) 18 6 9
Severe (3) 0 0
Overall 44 (19.4%) 14 (15.7%)
7% died. Although there were fewer females identified
with HIE, a greater n u m b e r of these were severely
handicapped or dead (13 of 39 vs 6 of 50, X2 = 9.855,
P = 0.043). There was a significant relationship between
the year of birth and outcome, with a decrease in the
occurrence of handicap from the first two complete years
(1975-1976) of the study to the last two years ( 1977-1978)
(16 of 28 vs 14 of 55, X'-' = 5.8, P = 0.027). Although over
100 variables from the antepartum and intrapartum
period were tested, none, alone or in combination, was
found to be significantly related to the asphyxia stage or
outcome.
Apgar score at one minute was not found to be
significantly related to the stage of asphyxia or outcome,
but there was a significant relationship between the
five-minute Apgar score and subsequent outcome, as
shown in Table III; the group with Apgar scores of 0 to 3
had significantly greater hzindicap than the other infants
(64% vs 30%, P < 0.001). Only four of the 14 infants with
the lowest five-minute Apgar scores had normal out-
comes.
O f the 17 infants who required intubation and pro-
longed ventilation in the delivery suite, ten were dead or
significantly handicapped, compared to 20 of 67 who
required other forms of resuscitation (X~-= 14.116,
P = 0.007).
Seizures were noted in 68% of infants in the first week
of life and correlated significantly with the encephalopa-
thy score (X2 = 9.46, P --- 0.024). Although there was no
significant relationship between the presence of seizures
and outcome (X~ = 4.34, P = 0.362), the age at the onset
of the seizures did relate significantly to outcome (Table
III), with 48% o f infants having seizures within 24 hours
being significantly handicapped compared to 24% of
infants whose seizures began after 24 hours (P = 0.049).
The presence of a low voltage or suppression pattern on
Hypoxic-ischemic encephalopathy ! 15
] Moderate Severe
handicap handicap Dead P value
1 3 5
0.001
11 10 1
8 5
0.04
3 0
0 0
5 1 0.001
1 6 5
12 (13.5%) 13 (14.6%) 6 (6.7%)
the initial E E G was also associated with an increased
probability of handicap (X~ = 14.2, P = 0.007).
Computerized axial tomography was available only for
the last six months of the study period; as a result, only 11
infants underwent this diagnostic procedure. There was
no relationship between the C T scan result and outcome.
No significant relationship was observed between the
presence of intracranial hemorrhage (diagnosed by a
bloody, nonclearing lumbar puncture, with hypoglycorr-
hachia, positive subdural taps, and C T scan when avail-
able, Table II) and either outcome or asphyxia stage. The
two children with skull fractures were severely handicap-
ped, but there was no relationship between the presence
of birth trauma (n = 17) or mode o f delivery and out-
come.
The relationship between the stage o f HIE and outcome
is shown in Table III. None o f the 25 children who had
encephalopathy Stage I died or had any significant
handicap, in contrast to all of those with Stage 1II
encephalopathy, who had moderate or severe handicap or
who died.
There was no significant relationship between neonatal
drug therapy and outcome.
Postmortem examinations were performed in three of
the four neonates dying of HIE, all of w h o m had evidence
of clinical brain death for 24 hours or more before death.
As a result, the neuropathology was unhelpful in defining
specific areas of damage.
DISCUSSION
In the present study, 31 of 89 term infants (35%) had
moderate to severe handicap or died as a result of their
HIE. This is somewhat less than the figure of 48% in the
study by Brown et al. 3 The incidence of significant HIE in
term infants in this hospital from September, 1974, to
December, 1978, was 3.6/1,000 live births of 37 weeks or
1 16 Finer et al.
greater, slightly greater than the 0.2% incidence for infants
of 37 weeks or greater recently reported by MacDonald et
al. 15 The outcome was not significantly different for
inborn children vs those transferred, and resulted in an
incidence of Severe handicap or death of 1.3/1000 live
births of 37 weeks or greater.
Although previous studies have demonstrated a poor
prognosis for infants with neonatal seizures," . . . . ~ in this
study, restricted to term infants with documented HIE,
there was no demonstrable relationship between the
presence of seizures and outcome. There was, however, a
significant relationship between the timing of the onset of
seizures and outcome with infants having earlier seizures
being at greater risk for an abnormal outcome (Table III).
The association between the presence of a low voltage
background on EEG and subsequent outcome is in
keeping with previous observations?. 5.1~ and with the
findings of Sokol et aP 7 that prolonged voltage depression
on fetal EEG bears a significant relationship to abnormal
neurobehavorial outcome. The incidence of intracranial
hemorrhages may have been significantly underestimated
because of the limited availability of CT scanning.
Sarnat and Sarnat ~ reported a group of term infants
with evidence of fetal distress and described three clinical
stages of neonatal postanoxic encephalopathy. They dem-
onstrated that infants who entered Stage III, who had
signs of Stage II for more than seven days, or whose EEG
failed to revert to normal either died or were found to
have significant neurologic impairment. Our results con-
firm the observations of Sarnat and Sarnat 1 in that there
was a demonstrable relationship between the stage of
encephalopathy (scored without EEG criteria) and out-
come. In their study, five of 21 infants had evidence of
Stage III encephalopathy and two died; the remaining
three were abnormal on examination at six to 11 months.
In our study, there were 12 infants who had Stage III
encephalopathy and all of these infants were either dead
or significantly handicapped.
There was a significant association between the five-
minute Apgar score and outcome, especially for the group
with Apgar score of 0 to 3, which is in keeping with the
initial observations of Drage and Berendes TM and of
Berendes? ~ More recently, Nelson and Ellenberg~~found
that an Apgar score of 0 to 3 at five minutes was an
ominous finding with 44% of patients dying, 5% of all
surviving infants showing evidence of disabling motor
deficits, and the relative risk for cerebral palsy in such
infants being 21 times normal.
Volpe ~ has stressed that the value of neonatal neuro-
logic examination is greatest when taken "in the context
of the neuropathologic disorder." In the present study, all
infants were 37 weeks or greater and felt to have hypoxic-
The Journal of Pediatrics
January 1981
ischemic encephalopathy on the basis of their perinatal
histories and clinical examinations. For this group of
infants, seizures per se appear to have no prognostic
value, but the five-minute Apgar score, need for vigorous
resuscitation, encephalopathy stage, time of onset of
seizures, sex, and background EEG pattern are all signif-
icantly related to outcome. It would appear that the use of
the Sarnat staging system without EEG does allow for
reasonably accurate prognostication regarding the proba-
bility of future handicap.
In spite of the fact that the maternal population was
highly abnormal, both in the presence of complications
and in the modes of delivery, there was no specific
correlation between any of the antepartum or intrapartum
variables examined and subsequent outcome. This may
be related to the small number of patients examined and
the lack of appropriate information, or may reflect the
fact that the resulting decrease in blood flow and/or
oxygen delivery to the fetus was more important than the
specific etiology. Mulligan et aF 2 were also unable to
demonstrate a relationship between outcome and prenatal
complications in their study of 65 survivors of neonatal
asphyxia.
Improved recognition and more aggressive therapy of
HIE postnatally may reduce the mortality and morbidity
of this condition.
The authors gratefully acknowledge the invaluable Contribu-
tion of the nurses and staff of the Perinatal-Neonatal Unit in the
Royal Alexandra Hospital. We also thank Dr. P. Etches and Dr.
A. Stewart for their thoughtful review and suggestions, and Mrs.
M. Snelling for her expert secretarial assistance in the prepara-
tion of this manuscript.
REFERENCES
1. Sarnat HB, and Sarnat MS: Neonatal encephalopathy
following fetal distress. A clinical and electroencephalo-
graphic study, Arch Neurol 33:696, 1976.
2. Volpe JJ: Observing the infant in the early hours after
asphyxia, in Gluck L, editor: Intrauterine asphyxia and the
developing fetal brain, Chicago, 1977, Yearbook Medical
Publishers, Inc., pp 262-283.
3. BrownJK, Purvis RJ, Forfar JO, and Cockbum F: Neuro-
logical aspects of perinatal asphysia, Dev Med Child Neurol
16:567, 1974.
4. Amiel-Tison C: Neurologic disorders in neonates associated
with abnormalities of pregnancy and birth, Curt Probl
Pediatr 3:1, 1973.
5. Brann AW Jr, and Dykes FD: The effects of intrauterine
asphyxia in the term neonate, Clin Perinatol 4:149, 1977.
6. Hon EH: An atlas of fetal heart rate patterns, New Haven,
1968, Harty Press.
7. Werner SS, Stockard JE, and Bickford RG: Atlas of
neonatal electroencephalography, New York, 1977, Raven
Press.
8. GeseUA: The first five years of life: A guide to the study of
Volume 98
Number 1
the pre-school child. Part One: early mental growth, New
York, 1940, Harper & Rowe Publishers, lnc, Chaps. 3
and 4.
9. lUingworth R: The development of the infant and young
child, normal and abnormal, ed 4, Edinburgh, 1970, E & S
Livingstone Ltd., Chap 9.
10. Mecham M J: Verbal language developmental scale, Circle
Pines, Minnesota, 1971, American Guidance Service.
11. Uzgiris IC, and Hunt J.McV.: Assessment in infancy--
Ordinal scales of psychological development, Urbana, 1975,
University of Illinois.
12. Stanford-Binet intelligence scale, in Terman LM, and
Merill, MA, editors: Boston, 1972, Houghton Misslin Com-
pany.
13. Berry KE: Developmental test of visual-motor integration,
in Administration and scoring manual, Chicago, 1967,
Follet Publishing Company, p 14.
14. Dubowitz LMS, Dubowitz V, and Goldberg C: Clinical
assessment of gestational age in the newborn infant,
J pEDIATR 71:1, 1970.
15. MacDonald HM, Mulligan JC, Allen AC, and Taylor PM:
Neonatal asphyxia. I. Relationship of obstetric and neonatal
H y p o x i c - i s c h e m i c encephalopathy 1 17
complications to neonatal mortality in 38,405 consecutive
deliveries, J PEDIATR 96:898, 1980.
16. Rose AL, and Lombroso CT: Neonatal seizure states. A
study of clinical, pathological, and electroencephalographic
features in 137 full-term babies with a longterm follow-up,
Pediatrics 45:404, 1970.
17. Sokol R, Rosen J, and Chik L: Fetal electroencephalo-
graphic monitoring related to infant outcome, Am J Obstet
Gynecol 127:329, 1977.
18. Drage JS, and Berendes HW: Apgar scores and outcome of
newborn, Pediatr Clin N A m e r 13:635, 1966.
19. Berendes HW: Brain damage in the fetus and newborn
from hypoxia, in James LS, Meyer RE, and Gaull GE.
editors: Columbus, Ohio, 1967, Ross Laboratories.
20. Nelson KB, and Ellenberg JH: Neonatal signs as predictors
of cerebral palsy, Pediatrics 64:225, 1979.
21. Volpe JJ: Value of the neurologic examination, Pediatrics
64:547, 1979.
22. Mulligan JC, Painter M J, O'Donoghue PA, MacDonald
HM, Allen AC, and Taylor PM: Neonatal asphyxia. II.
Neonatal mortality and long-term sequelae, J PEDIAa'R
96:903, 1980.