Mortality
and
Morbidity
in Infants
Less
Than
1 ,001
Grams
Birth
Weight
John M. Driscou, Jr, MD, Yvonne T. Driscoll, MD, Mary E. Steir, MS
Raymond I. Stark, MD, Barbara C. Dangman, MD, Alicia Perez, MD,
Jen-Tien Wung, MD, and Paula Kritz, PNP
From the Department of Pediatrics, College of Physicians & Surgeons, Columbia University and Babies Hospital, New York
ABSTRACT. A prospective study of 54 infants with birth
weights of 1,000 gm or less was conducted over a period
of two years. Of the 26 infants who survived, 24 weighed
between 750 and 1,000 gm; two infants died after
dis-charge and one was lost to follow-up, leaving 23 in whom
serial observations were made over 18 months to 3 years
of age. The incidence of neurologic deficit in these infants was 17% and of intellectual deficit, 13%. Of the four who were abnormal neurologically, two had spastic quadripa-resin, one static encephalopathy, and one hydrocephalus secondary to intraventricular hemorrhage. The three with intellectual deficit had a developmental quotient <85. Of
the perinatal factors examined, only birth asphyxia cor-related significantly with both neonatal mortality and
subsequent morbidity. Six (26%) of the surviving infants
had mild, nonblinding retrolental fibroplasia; only one of them had a significant refractive error that required
cor-rective lenses for vision. Sepsis was a significant
contrib-utor to neonatal mortality in ten of 28 infants who died,
but was detected in only one survivor. Although the
prognosis for the infant weighing 1,000 gm or less at delivery has improved significantly, there is promise for still further improvement by reducing perinatal asphyxia.
Pediatrics 69:21-26, 1982; very low birth weight,
mortal-it)’, morbidity, retrolental fibroplasia, sepsis.
Infants with birth weights of 1,000 gm or less
have been at great risk for neurologic, intellectual,
and behavioral sequelae.’3 A little more than a
decade ago only 10% of infants in this weight
cate-gory survived.4 Recently, because of better obstetric
and pediatric care, the survival of the very
low-birth-weight infant has improved.7 With this
im-provement, the quality of life for surviving infants
has become an important issue. Until now, limited
Received for publication Aug 8, 1980; accepted April 24, 1981. Reprint requests to (J.M.D.) Department of Pediatrics, Babies Hospital, Box 13, 3959 Broadway, New York, NY 10032.
PEDIATRICS (ISSN 0031 4005). Copyright © 1982 by the
American Academy of Pediatrics.
information describing the intellectual, physical,
and motor development of recent survivors has
been available.8’9
PATIENTS AND METHODS
Between Jan 1, 1977, and Dec 31, 1978, 54 infants
weighing 1,000 gm or less were admitted to the
Neonatal Intensive Care Unit of Babies Hospital;
30 were born in the hospital (inborn) and 24 were
referred to the tertiary center by hospitals involved
in a perinatal network (outborn). Of the 26 infants
(48%) who survived, two died after discharge from
the hospital, 23 were available for follow-up over an
18- to 36-month period, and one was lost to
follow-up.
On admission to the unit, all infants entered a
prospective study oflongitudinal outcome. The
rec-ords of all patients were reviewed, and pertinent
antenatal, intrapartum, and neonatal information
was collected. Informed consent for participation in
the study was signed by families before discharge
from the hospital.
Gestational age was determined by a combination of mother’s expected date of confinement, physical,
neurologic, and ocular assessment, and in certain
infants by motor nerve conduction velocity. All
received care in Servo-controlled incubators that
maintained skin temperatures at 36.5 C. When
nec-essary, umbilical arterial catheters were placed;
in-fusions through peripheral veins were started in
patients who did not require arterial
catheteriza-tion. Oral feedings with breast milk or synthetic
formulas were started as early as possible by
naso-gastric tube. Until infants were able to tolerate
enteral feedings, parenteral solutions with glucose,
5 to 12.5 mg/i#{174} ml, and amino acids, 2.5 mg/100
ml, were infused. Intralipid was also given to
ml. The total fluid administered to each infant on
the first day was 80 mi/kg/day and the volume was
gradually increased as tolerated to 150 mi/kg/day
by a combination of oral and parenteral fluids.
Ambient oxygen concentration was monitored
continuously. Transcutaneous Po2 was recorded
constantly and the arterial oxygen tension was
mea-sured directly at least every four hours. When it
was not possible to maintain an adequate Pao2 of
between 50 and 60 mm Hg, continuous distending
pressure (CDP) by nasal prongs was started. In
infants requiring mechanical ventilation for apnea,
inadequate arterial oxygen tension (Pao2 < 50 mm
Hg in 60% oxygen), or hypercarbia (Pco2 > 60 mm
Hg) the Bennett respirator was used with a
naso-tracheal tube. Apnea not responsive to stimulation
was treated with a combination of theophylline (5
mg/kg/day) and CDP by nasal prongs. Metabolic
acidosis was corrected with dilute solutions of
so-dium bicarbonate when the base deficit was more
than 10 mEq/liter.
Glucose levels were monitored by using
Dextro-stix and measured directly by the laboratory when
a low level was found by Dextostix. Biirubin levels
were determined in all infants on admission and
followed at least twice a day until the level had
peaked. Prophylactic phototherapy was used in
most patients in an attempt to avoid exchange
transfusion. (Exchange transfusion was performed
if biirubin levels reached 7 to 10 mg/i#{174} ml and if
the birth weight and birth history indicated that it
was needed.) Calcium and electrolyte levels were
determined in all babies and parenteral calcium
gluconate solution was given when calcium levels
were less than 7.5 mg/i#{174} ml.
Investigation for possible sepsis and/or
meningi-tis resulted in most patients having at least one
blood culture and lumbar puncture. Ampicillin and
gentamicin were usually given when antibiotics
were used for proved or suspected sepsis.
Antibiot-ics were not used prophylactically if umbilical
yes-sels were catheterized or if mechanical ventilation
was needed. In every infant, nasopharyngeal
cul-tures were used to study bacterial colonization to
determine whether there was an increased risk for
nf’#{176}”
Indirect ophthalmoscopy with cycloplegic
refrac-tion was performed by the same ophthalmologist
on all infants after oxygen therapy had been
discon-tinued. During their hospitalization, infants with
abnormal initial examinations were seen every four
to six weeks until the retinal changes had stabilized.
The frequency of outpatient eye examinations was
based on findings present at the last examination in
the hospital. All outpatient examinations were done
with cycloplegic refraction.
Parents and grandparents had unrestricted
visit-ing privileges and were encouraged to handle their
infants as early and as frequently as possible.
After discharge infants were seen at 6, 12, and 18
months of age and annually thereafter. All
neuro-developmental assessments and the Bayley Scales
of Infant Development’2 were corrected for
gesta-tional age. At the time of evaluation, children were
classified either as neurologically normal, suspect,
or abnormal after correction for gestational age.
The physician assessing the neurologic status of
these infants had not been involved in their nursery
care and was unaware of the details of the clinical
course. The psychologist had no knowledge of the
neonatal history before administration of the
Bay-ley Scales. Only infants with evidence of retrolental
fibroplasia (RLF) at discharge were followed by an
ophthalmologist. Hearing was tested at each visit
using clinical assessment.
Results have been analyzed by x2 test to compare
frequencies and Student’s t test to compare means
of normally distributed data. Differences reported
as significant are P < .05.
RESULTS
Mortality
Twenty-eight infants died, 22 during the first 28
days and six thereafter for a neonatal mortality of
41% and infant mortality of 15%. The two infants
who died after discharge from the unit have been
included in these rates. The combined neonatal and
infant mortality was 56%. Excluding the six infants
who died after 28 days, half died within the first 48
hours. No differences in the time of death were
noted between the inborn and outborn infants. The
primary causes of death based on postmortem
ex-aminations in 18 infants and clinical assessment in
ten infants are presented in Table 1.
Intraventric-ular hemorrhage (IVH) and hyaline membrane
dis-ease (HMD) accounted for 17 deaths, 16 of which
occurred during the first week of life.
Bronchopul-monary dysplasia (BPD), necrotizing enterocolitis
TABLE 1. Causes of Death
Postmortem Examination
Clinical Diagnosis Without Postmortem Examination (No.)
7 Intraventncular hemorrhage- (6)
2 Hyaline membrane disease- (2)
2 Bronchopulmonary dysplasia- (2)
2 Necrotizing enterocolitis
2 Sepsis
1 Disseminated intravascular coagulation
1 Oligohydramnios syndrome
1 CNS shunt infection
Died (28)
762 ± 126
27.5 ± 1.8
2.6 ± 1.6
5.1 ± 2.6
95.4 ± 1.5 P <.0001 <.01 <.0001 <.05 <.0001 Lived (26)
910.0 ± 98.8
28.3 ± 1.5
5.3 ± 2.6
6.75 ± 2.6
97.7 ± 1.0
3/26 1/26 8/26 11/26 12/26 1/26 18/28 <.0001
(NEC), and sepsis caused eight deaths, all of which
occurred after the seventh day of life. Deaths in the
remaining three infants were from disseminated
intravascular coagulation (DIC), oligohydraminios
syndrome, and CNS shunt infection, respectively.
The infant with shunt infection died at 67 days of
age. There were no cases of kernicterus documented
at autopsy.
Eleven factors differed between those who lived
and died (Table 2). Surviving infants had higher
birth weights, higher one minute Apgar scores,
higher temperatures on admission to the nursery,
and a lower incidence of clinical impression of CNS
hemorrhage, severe apnea, and the need for
me-chanical ventilation (P < .0001). Younger
gesta-tional age, low five minute Apgar score, presence of
HMD, sepsis, and seizures were significantly related
to mortality (P < .05). The indications for mechan-ical ventilation were the same for all weight groups
and the correlation between mortality and
mechan-ical ventilation reflects the more serious ifiness in
these infants, not their low weight. Cessation of
heroic therapy was based on the infant’s response
to treatment. Twenty-one infants (40%) had an IVH
by clinical assessment which was verified by
com-puted tomography (CT) scan in two; 18/21 died and
in 1 1 of these the hemorrhage was documented at
necropsy. The low frequency of documented IVH
during this period reflects the relative unavailability of CT scanning in our institution.
Several factors that correlated with mortality in
other series5”35 were examined. These included
place of birth, route of delivery, sex, intrauterine
growth, anemia, hypotension on admission, or
hy-pernatremia’6 during the course in the intensive
care unit. In our series none of these distinguished
between the infants who lived and died.
The incidence of cesarean section was higher in
the infants who died (12/28 vs 6/26), but these
differences did not reach statistical significance (P
= .1). The frequency of intrauterine growth
retar-dation plotted on the Lubchenco grid’7 was similar
in both groups with no difference in survival in the
growth-retarded infants. Furthermore, there was no
correlation between the occurrence of
hypernatre-mia and mortality nor between infants who were
inborn or outborn and mortality.
Neonatal Morbidity
Intrauterine growth retardation was noted in 21/
54 (38%) infants and was distributed evenly among
inborn and outborn infants and between the
surviv-ing and dying infants. Respiratory distress
syn-drome (RDS), diagnosed by clinical symptoms, a
reticulogranular pattern on chest roentgenogram,
and an oxygen requirement exceeding 72 hours,
occurred in 35/54 infants (65%). Thirteen of the 35
(37%) survived; mechanical ventilation was required
by 27, of whom seven (26%) survived. An additional
12 infants were ventilated for apnea and of these,
four (33%) survived.
Eight of 26 (30%) surviving infants had some form
of chronic lung disease, either BPD or chronic
pulmonary insufficiency of prematurity’8 diagnosed
by radiologic changes and clinical course during
their period of intensive care. All of these infants
had some abnormalities in respiratory symptoms at
discharge, but none was oxygen dependent. An
ad-ditional five infants who died during their
hospital-ization had BPD.
Patent ductus arteriosus (PDA) was diagnosed in
21/48 (44%) infants; six were excluded because of
early death. Diagnosis was based on a combination
of clinical history, chest roentgenogram, and
echo-cardiogram. Echocardiograms were not routinely
obtained for all patients, but were performed in
those infants suspected of having PDA. Thirteen of
the 21 infants survived (62%). PDA was complicated
________ ______
by congestive heart failure in 13, only one of whomresponded to medical treatment; the remaining 12
(92%) required ligation of the ductus and eight
(75%) survived. In the four infants who died, the
cause of death was not related to surgery, but to
chronic lung disease and sepsis. Indomethacin was
infrequently used in the nursery during the period
of this study and was not given to any of the infants in this series.
Sepsis was diagnosed in 11/54 infants. Only one
28/28 <:#{174} of the infants with documented sepsis survived; 10/
28 (36%) of the infants who died had a positive
20/28 <.05 blood culture antemortem. The most common
or-ganisms were Staphylococcus aureus (three cases),
9/28 <.05 Pseudomonas (two cases), and Haemophilus
influ-TABLE 2. Differences Between Infants Who Lived and
Died
Birth weight (gui)
Gestational age (wk) Apgar score
1-nun
5-mm
Temperature on
enzae (two cases). The remaining organisms that
were isolated included Cornyebacterium, Listeria,
Escherichia coli, and Candida. There was no sta-tistically significant difference in duration of
rup-ture of membranes or place of birth between the
infants with sepsis and those with negative cultures.
Two of ii infants with positive blood cultures had
simultaneously positive CSF cultures (S aureus and
E coli).
NEC diagnosed by clinical history, pneumatosis
on roentgenogram, or at surgery or postmortem
examination, occurred in 4/54 (7%) infants. RLF
was diagnosed in 6/26 (23%) surviving infants; four
infants had grade II disease and one infant each
had grade I and grade III disease. Only the infants
with evidence of RLF were examined by an
ophthal-mologist after discharge and their last examination
occurred between the ages of 8 and 18 months,
depending on the grade of RLF. Three of the six
infants with RLF required no greater than 30%
oxygen at any time, but all had apnea requiring
Ambu bag ventilation with the gas mixture in the
bag being no greater than 30%. At present, only the
infant with grade III disease requires corrective
lenses for vision.
Outcome
Among the 26 survivors, there were two deaths,
one due to sudden infant death syndrome (SIDS)
and the second due to BPD. At the time of
dis-charge, neurologic assessment was normal in the
infant who was a victim of SIDS and suspect in the
infant with chronic lung disease. Neither infant had
been evaluated in the follow-up clinic before death.
A third infant has had no follow-up evaluation. The
23 remaining have been followed prospectively for
their neurologic, intellectual, and somatic develop-ment.
At the most recent examination, neurologic
de-velopment was normal in 19/23 infants (83%) whose
ages range from 8 months to 3 years. Twenty-two
infants were seen between ages 18 and 36 months.
The remaining infant was not accessible for
follow-up, but was completely normal neurologically at 8
months of age. The neurologic sequelae in the four
abnormal infants are severe and include spastic
quadriparesis (two infants), severe static
encepha-lopathy (one infant), and hydrocephalus secondary
to an IVH (one infant). The one- and five-minute
Apgar scores were the only factors that significantly
differed between normal and abnormal children
(Table 3). There were a number offactors that were
consistently present in the infants who were
neu-rologically abnormal. These were appropriate
growth, inborn delivery, hypotension (less than 40
mm Hg mean systolic pressure on admission),
hy-TABLE 3. Relationship of Neurologic Outcome and
Apgar Score
Apgar Score Neurologic Status
Normal Abnormal P
i-miri 6.2 ± 2.3 2.2 ± 2.5 <.006
5-mm 7.6 ± 2.1 3.8 ± 2.8 <.005
pernatremia (greater than 145 mEq/liter), and
hy-poxia (less than 40 mm Hg during the first 24
hours). It is noteworthy that the occurrence of these factors was not significantly different in the normal
survivors. The mean peak bilirubin levels of the
neurologically normal and abnormal children were
nearly identical (8.4 ± 1.7 mg/i#{174} ml vs 8.4 ± 1.5 mg/i#{174} ml, respectively). Three of four abnormal
children and six of 19 normal children required
mechanical ventilation. Four of the seven infants
who were abnormal either neurologically or by
Bay-ley scores required ventilation for either severe
apnea or birth asphyxia. Sepsis was not documented in either group.
Mental performance was assessed in 16/19
neu-rologically normal infants. The mean age at testing
was 17 ± 3.0 months and the mean mental
devel-opmental index (MDI) was 98.2 ± 19. There were
three children whose MDI was <85 in this group.
Two of these children had intrauterine growth
re-tardation and a complicated perinatal history; one
was born to a mother who consumed a quart of
vodka a day throughout her pregnancy; the other
had NEC with a prolonged period of central and
peripheral alimentation. No apparent cause for the
growth retardation could be ascertained in the third
infant. Although the children in the study were
from a mixed socioeconomic background, all three
infants with MDI <85 were from the lower
socio-economic group. Two of three infants with low
scores were also inborn. None of the four infants
who were neurologically abnormal could be
satis-factorily tested with the Bayley Scales of Infant
Development and will be tested at a later date.
Interestingly, all four infants with neurologic
ab-normalities were from higher socioeconomic groups.
Somatic growth for height and weight was normal
when corrected for gestational age in 18/19 infants
who were neurologically normal. The remaining
child was proportionately small in all three
param-eters and had parents with small stature. Five
chil-dren with normal height and weight measurements
had a head circumference that was less than the
third percentile. However, in all these children, the
head circumference was <2 SD different from the
height and weight measurement and four of them
had intrauterine growth retardation; the MDI was
infants who were small for gestational age (SGA)
achieved normal proportionate height and weight.
DISCUSSION
Prior to 1975, the neonatal survival rate for
in-fants 1,000 gm or less was iO%. The decrease in
mortality in the present report is similar to figures
reported from other centers.5’7’9 In our series no
infant <700 gm survived while 89% of infants
be-tween 901 and 1,000 gm survived. Similar trends in
the weight distribution of survivors are also appar-ent in the experience of others.6”9 Interestingly, the
absolute number of infants in this weight group
admitted to our institution more than doubled when
compared with statistics for 1970 to 1971. During
the period of this study, infants 1,000 gm or less
constituted 5.4% of all admissions compared to 3.0%
of admissions during the earlier period. Both the
increasing percentage of intensive care unit
admis-sions for this weight group and the improved
neo-natal survival reflect a more aggressive approach to
treatment in the perinatal period.
IVH and HMD continued to be the leading causes
of death within the first 7 days of life; BPD and
NEC were the most significant contributors to
deaths that occurred after the first week oflife. Our
findings at post mortem confirm a recent study#{176}
that documented the inaccuracy of clinical
assess-ment in detection of IVH. However, institutional
limitations at the time prohibited CT scans in all
infants, leaving only clinical assessment as an
indi-cator of approximate incidence of this disorder. As
in prior studies, birth asphyxia significantly
contrib-uted to neonatal mortality and was also directly
related to later neurologic morbidity. However,
con-trary to prior reports,5”3 we observed no differences
in mortality or neurologic outcome related to
method of delivery.
The improved survival in our series has not been
associated with a high neurologic morbidity as
noted in earlier reports.’3 In fact, the longitudinal
outcome compares favorably with the results
ob-tamed in infants with birth weights between 1,000
and 1,500 gm. Although our incidence of disabling
neurologic disorders is lower than in earlier reports
for this weight group, the total longitudinal compli-cation rate including intellectual impairment is stifi 30%.
Although the incidence of RLF’4’’ is 23% in
sur-viving infants, the disease was transient in 5: none
are blind, and only one infant with grade III disease is significantly myopic. The other infants with grade
I and grade II RLF are not considered to have
significant refractive errors. As all infants without
RLF were not followed by an ophthalmologist, the
true incidence of refractive error in this group is
unknown. It is interesting to note that three of six
infants with RLF received no more than 30%
am-bient oxygen with one infant receiving 30% oxygen
for less than one hour. All six infants had apnea
requiring hand ventilation with a resuscitation bag
for treatment. The concentration of oxygen
admin-istered during treatment was the same as that given
during spontaneous breathing. The retinal
imma-turity ofthese infants may well place them at higher
risk for the development of RLF.
A PDA was diagnosed in 16/26 (62%) surviving
infants, an incidence higher than previously
re-ported in this weight group.’4 Likewise, the
inci-dence of congestive heart failure and the failure to
respond to medical treatment was higher in our
series. The time of this study preceded a controlled
clinical trial to clarify the role of indomethacin in
the management of infants with a PDA. if
prelimi-nary evidence that indomethacin may be less
effec-tive in this weight group is true,22 the lack of surgical
morbidity related to operative closure of the duct
in our patients is encouraging.
The incidence of intrauterine growth retardation
in our population was similar to that in a prior
report.’4 However, contrary to other studies,8’ the
infants who were growth retarded did not have a
higher survival rate and eight ofnine surviving SGA
infants had somatic growth that achieved the third
to 90th percentile for height and weight,’ as well as
Bayley scores (100 ± 15) similar to the
appropri-ately grown infant (98 ± 23). It is encouraging to
note that the somatic growth and the neurologic
and intellectual development of the SGA infant of
<1,000 gui birth weight compares favorably with
his appropriate for gestational age (AGA)
counter-part in our series.
Compared with earlier reports,6 the absence of
kernicterus in the 18 postmortem examinations and
the absence of athetosis and deafness in survivors
is surprising in these sick, very low-birth-weight
infants. The lack of anatomic or clinical evidence of
bilirubin toxicity in this group may reflect the use
of prophylactic phototherapy and our use of a lower
(10 mg/i#{174} ml) level ofbilirubin for exchange
trans-fusion.
Although only nine of the surviving infants
re-qui.red mechanical ventilation, the abnormal
neu-rologic outcome in three of these children is similar to findings reported previously.27 The infants
weigh-ing <1,000 gui who required mechanical ventilation
remain at significant risk for neurologic sequelae.
Compared with previous reports, the relatively
low incidence of neurologic and intellectual
morbid-ity in these very low-birth-weight infants is
encour-aging. In light of the increasing number of small
infants who are surviving, and the significant
a more aggressive approach to antepartum and
intrapartum monitoring of these infants would seem
important. If further improvement in neurologic
and intellectual outcome is to be achieved, it will
be through prevention of perinatal asphyxia.
ACKNOWLEDGMENTS
This work was supported in part by National Institutes of Health grant HL 14218, contract HR 5-2948.
We are grateful to Dr L. Stanley James for his
assist-ance. We thank Wilda Santana who coordinates the
fol-low-up program and Lybia Marsi for typing the
manu-script.
REFERENCES
1. Drillien CM: The incidence ofmental and physical handicaps in school age children of very low birth weight. Pediatrics
27:452, 1961
2. Finch RO, Graven HL, Engel RR: Neurological status of survivors of neonatal respiratory distress syndrome. J Pe-diatr 73:3, 1968
3. Kantero RL, Frisk M, Holmstrom G: Small prematures at 6-7 years of age. Ann Paediatr Fenn 11:175, 1965
4. Behrman RE, Babson GS, Lessel R: Fetal and neonatal mortality in white middle class infants, mortality risks by gestational age and weight. Am J Dis Child 121:486, 1971 5. Bowes WA: Results of the intensive perinatal management
of very low birth weight infants (501-1500 grams), in Pro-ceedings of the Fifth Study Group of Royal College of
Obstretricians and Gynecologists. 1977, p 331
6. Schechner S: For the 1980’s. How small is too small? Clin
Perinatol 7:135, 1980
7. Haessleen HC, Goodlin RC: Delivery of the tiny newborn. Am J Obstet Gynecol 134:192, 1979
8. Stewart AL, Turcan DM, Rawlings G, et al: Prognosis for infants weighing 1000 grams or less at birth. Arch Dis Child
52:97, 1977
9. Hack M, Fanaroff AA, Merkatz JR: The low birth weight infant. N Engl J Med 301:1162, 1979
10. Sprunt K, Leidy G, Redman W: Abnormal colonization of neonates in an intensive care unit: means of identifying neonates at risk of infection. Pediatr Res 12:998, 1978 11. Sprunt K, Leidy G, Redman W: Abnormal colonization of
neonates in an intensive care unit: conversion to normal colonization by pharyngeal implantation of alpha hemolytic Streptococcus strain 215. Pediatr Res 14:308, 1980 12. Bayley N: Manual for the Bayley Scales of Infant
Devel-opment. New York, The Psychological Corporation, 1969 13. Stewart A: Follow-up of pre-term infants, in Proceedings of
the Fifth Study Group ofthe Royal College of Obstreticians and Gynecologists. 1977, p 372
14. Pape KE, Buncie RJ, Ashby S et al: The status at two years of low birth weight infants born in 1974 with birth weights of less than 1001 grams. J Pediatr 92:253, 1978
15. Alden ER, Mandelkorn T, Woodrum DE, et al: Morbidity and mortality of infants weighing less than 1000 grams in an intensive care nursery. J Pediatr 50:40, 1972
16. Simmons M, Adcock E, Bard H, et al: Hypernatremia and intracranial hemorrhage in neonates. N Engl J Med 291:6, 1979
17. Lubchenco LO, Hansman C, Boyd E: Intrauterine growth in length and head circumference as estimated from live birth and gestational age from 26-42 weeks. Pediatrics 37:403, 1966
18. Wung JT, Koons AH, Driscoll JM, Jr, et al: Changing incidence of bronchopulmonary dysplasia. J Pediatr 95:845, 1979
19. Roberton NRC: Intensive care and the very low birth weight infant. Lancet 1:362, 1979
20. Silverboard G, Horder MH, Ahmann PA, et al: Reliability of ultrasound in diagnosis of intracerebral hemorrhage and posthemorrhagic hydrocephalus: comparison with computed tomography. Pediatrics 66:507, 1980
21. Fitzhardinge PM: Early growth and development in low birth weight infants following treatment in an intensive care nursery. Pediatrics 56:162, 1975
22. Friedman WF, Kurlinski J, Jacob J, et al: The inhibition of prostaglandin and prostacyclin synthesis in the clinical man-agement of patent ductus arteriosus. Semin Perinatol 4:125, 1980
23. Jones RAK, Cummins M, Davies PA: Infants of very low birth weight. Lancet 2:1332, 1979
24. Fitzhardinge PM: Intrauterine growth retardation: An added risk to the preterm infant, in International Neonatal Inten-sive Care Conference. Durham, NH, August 1977 25. Ackerman BD, Dyer GY, Leydort MM: Hyperbffirubinemia
in small premature infants. Pediatrics 45:906, 1970 26. Ackerman BD, Dyer GY, Leydort MM: Hyperbilirubinemia
in small premature infants. Pediatrics 45:918, 1970 27. Fitzhardinge PM, Pape T, Arstekaitis M, et al: Mechanical
