93% WWIWIIH“HIM!WHIIWWHIWIIIWI

     

HTHS

THES'S
/

’7 ryf‘x

r 5“ “V”

This is to certify that the

thesis entitled

THE ROLE OF OBSTETRIC FACTORS IN NEURODEVELOPMENTAL
OUTCOMES AND NEONATAL DEATH IN LOW BIRTH WEIGHT BABIES

presented by

HONG QIU

has been accepted towards fulfillment
of the requirements for

JASIELdegree in W0 G Y

KW

Major professor

Date // I/Z} /GO

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State

University

 

 

PLACE lN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

JULO fl 2 IZEDEfi

 

 

“3439057

 

 

 

 

 

 

 

 

 

 

 

 

woo woman-Drum.“

 

 

THE ROLE OF OBSTETRIC FACTORS IN NEURODEVELOPMENTAL
OUTCOMES AND NEONATAL DEATH IN LOW BIRTH WEIGHT BABIES

By

Hong Qiu

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Epidemiology

2000

ABSTRACT

THE ROLE OF OBSTETRIC FACTORS IN NEURODEVELOPMENTAL
OUTCOMES AND NEONATAL DEATH IN LOW BIRTH WEIGHT BABIES

BY
Hong Qiu

The objective of this study was to evaluate the relationship between labor
and delivery characteristics and neonatal abnormalities and childhood handicap
(neonatal death, neonatal brain damage, and disabling cerebral palsy (CP)).
The analysis was based on a cohort study of 1,105 infants with birth weights of
500-2,000 grams. Neonatal brain damage was assessed by ultrasound
screening. CP and disabling CP were assessed at two years of age (corrected
for gestational age) follow up. In unadjusted cell stratified analysis, presence of
active labor was directly related to all outcomes evaluated. When active labor is
combined with other risk factors (eg. amnionitis, placenta abruptio), it is
associated with higher risk of adverse outcomes. After controlling for possible
confounders, active labor still had a significant association with neonatal brain
damage and death (OR=1.4 -2.0). Active labor was not associated with DCP; the
effects of iatrogenic labor compared to spontaneous preterm labor on outcomes
were not different. The relationships of active labor and brain damage were
modified by intrauterine growth. The association of active labor with adverse
outcomes is increased in SGA babies as compared to normal for gestation
babies. The effect of active labor is also increased in other infant compromise.
The clinical significance of active labor is its effect on mortality and short-term

neurological damage in LBW infants, but not on childhood handicaps.

ACKNOWLEDGMENTS

I am grateful to many people who assisted me during my study in
Epidemiology.

I would like to express my most sincere gratitude to my advisor, Dr. Nigel
Paneth for giving me this wonderful opportunity to work on his project and his
guidance, enthusiasm and support in my graduate study. I also thank my
committee members, Dr. Michael Collins and Dr. Jack .M. Lorenz, for their kind
help and their insightful ideas of my research project. Their knowledge and
encouragement were vital resources of this thesis and are greatly appreciated. I
also thank Mrs. Madeleine Lenski for her enthusiastic help in both my English
improvement and daily life. I would like to give my thanks to Dr. Joseph
Gardiner, Dr. Wenjiang Fu and Mrs. Jenny wang for their assistance on
improving my statistical skills and their friendships. Thanks to Dr. Claudia
Holzman, James Jetton, Lora McAdams and all others in the department for
their support and kindness.

Finally, sincerest thanks to my parents and my husband Ying Wang for

their strongest support and love.

In.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................... v

LIST OF FIGURE ............................................................................................... vi

LIST OF ABBREVIATIONS ................................................................................. vii

CHAPTER 1

INTRODUCTION ................................................................................................... 1
The problem of low birthweight babies ....................................................... 1
Cerebral palsy and Neonatal brain damage in LBW ................................... 2
The effect of obstetric factors on fetus ........................................................ 5
Previous studies .......................................................................................... 8

CHAPTER 2

MATERIALS AND METHODS ............................................................................ 14
Study Population ...................................................................................... 14
Outcome assessment .............................................................................. 15
Exposure assessment .............................................................................. 17
Other birth characteristics ........................................................................ 20
Statistical analysis. .................................................................................. 20

CHAPTER 3

RESULTS ........................................................................................................... 22
Sample characteristics ............................................................................. 22
Risks of outcomes .................................................................................... 26
Univariate and multivariate analysis ........................................................ 32

CHAPTER 4

DISCUSSION ...................................................................................................... 39

REFERENCES ................................................................................................. 45

LIST OF TABLES

Table 1.1 Previous studies .................................................................................... 9
Table 1.1 Previous studies (cont’d) .................................................................... 10
Table 2.1: Criteria for the labor and delivery exposure ....................................... 18
Table 2.2: Criteria for the cause for preterm labor ............................................. 19
Table 3.1: Characteristics of cohort .................................................................... 24
Table 3.2: Unadjusted cell specific risk (°/o) of NH ............................................ 27
Table 3.3: Unadjusted cell specific risk (°/o) of PELVE ....................................... 28
Table 3.4: Unadjusted cell specific risk (°/o) of Neonatal death ........................... 29
Table 3.5: Unadjusted cell specific risk (%) of DCP ............................................ 30
Table 3.6: Adverse outcomes in relation with labor & delivery comparing by any
labor and spontaneous labor only ...................................................... 32
Table 3.7 Logistic regression for GM/IVH ........................................................... 35
Table 3.8: Logistic regression for PEL/VE .......................................................... 36
Table 3.9: Logistic regression for neonatal death ............................................... 37
Table 3.10: Logistic regression for DCP ............................................................. 38

LIST OF FIGURE

Figure 3.1 Distributions of exposure and outcomes ........................................... 23

vi

LIST OF ABBREVIATIONS

AGA-appropriate for gestational age

CLD-chronic lung disease

CP-cerebral palsy

C/S-cesarean section

DCP-disabling cerebral palsy

FGR-fetal growth ratio

GM/IVH- Germinal matrix/intraventricular hemorrhage
IUGR-intrauterine growth retardation

LBW — low birth weight

NICU-neonatal intensive care unit

PEL/VE- Parenchymal echodensities/lucencies Iventricular enlargement
PROM-preterm premature rupture of membrane
SGA-small for gestation age

VLBW-very low birth weight

vii

CHAPTER 1

INTRODUCTION

The problem of low birthweight babies:

Infants born at low birth weight contribute substantially to three national
burdens—infant mortality, childhood handicap and medical expenditure. In the
US. in 1997, 65% of all infant deaths occurred in the 7.5% of LBW
infants(<2,5009), and 51% of all infant deaths occurred in the 1.4% of VLBW
infants(<1,5009). Compared with infants of normal birthweight, infants weighing
1,500 to 2,499 9 have a 5 times higher risk of dying in the first year of life and
VLBW infants experience a 92-fold excess risk[1], while survivors experience a
twenty-five fold excess risk of cerebral palsy [2]. Low birth weight babies also are
at higher risk for mental retardation [3], blindness [4], deafness [5], seizure
disorders [6], Ieaming disabilities [7] and chronic lung disease [8]. A moderate or
severe disability or handicap is found in 25% of VLBW infants [9, 10] and, if
serious school difficulties are included, 40-50% of VLBW survivors are disabled
[11].

The poor international infant mortality ranking of the US is due principally
to the high LBW rate. This rate has actually worsened in the US. in the past two
decades [12, 13]. Between 1981 and 1997, the birth prevalence of VLBW rose by
22% in white infants and by 21% in black infants [14]. Unfortunately, expanded
prenatal care, social support programs, antibiotic treatment and other efforts

have failed to lower the risk of premature birth [15-17]. In the past 20 years,

neonatal intensive care has been in the forefront of medical advances, and
survival of low birth weight and preterm babies increased significantly. A
comparison of weight-specific mortality rate in a 23-year interval (1960-1983) in
the US. showed an astonishing decrease in mortality from 99% to 54% in VLBW
singleton infants [18]. By 1996, infant mortality for VLBW had dropped to
34%[19]. This continuing increase in the number of surviving small babies draws
attention to questions concerning the quality of life of new survivors.
Unfortunately, the prevalence of disabilities has not changed among small babies
now experiencing increasing survival. An apparently unavoidable side effect of
the increasing success of newborn intensive care is a steadily rising prevalence

of childhood neurodevelopmental handicap [20, 21].

Cerebral palsy and Neonatal brain damage in I__BW:

Cerebral palsy (CP), one of the major developmental disabilities of
childhood, is a group of non-progressive impairments of motor function
originating from the central nervous system dating back to events in the prenatal
or perinatal period. As birth weight and gestational age decrease, the prevalence
of CP increases significantly. The increase begins clearly at <35 weeks
gestational age. As a non fatal yet incurable disease, cerebral palsy is one of the
most costly handicaps among birth defects. It produces a huge financial and
psychological burden to affected families and society. The etiology of CP is not
clearly established. The etiologic events most usually cited are ischemia and /or

asphyxia, brain white matter damage and perinatal infection.

Neonatal brain damage, a very common finding in preterm infants, has
been found to be a strong predictor of neurological impairment. In 1979, Pape
and colleagues [22] demonstrated that brain hemorrhage in infants could be
imaged successfully by the use of ultrasound beams aimed through the anterior
fontanelle. Several years later, Levene et al [23] reported that other brain lesions,
e.g. periventricular leukomalacia, could be diagnosed on ultrasound. These
findings opened up the possibility of screening all neonates because the
ultrasonographic procedure has no known side effects, is inexpensive, and can
be performed in the nursery at the bedside. In the following decades, cranial
ultrasonography became a standard intensive care nursery procedure throughout
the US and Europe. The reliability of this technique has been confirmed in our
cohon[24]

Germinal matrix/intraventricular hemorrhage (GM/IVH) is the most
frequent and clinically significant brain damage found in preterm infants,
occurring in about 40% of very low birth weight infants [22, 23]. The clinical
significance of GM/IVH lies principally in its contribution to mortality. Many LBW
infant survivors GM/IVH without ventricular enlargement of parenchymal
hemorrhage are found to be asymptomatic and show no major neurological
abnormality or developmental delay [25-28]. The brain lesions with the most
unfavorable prognoses for major handicap in LBW children are parenchymal
echodensities/echolucent lesions or ventricular enlargement (PEL/VE) [25, 27-
29]. These lesions have been found to be associated with infarction or necrosis
of cerebral white matter (sometimes referred to as periventricular leukomalacia)

[30]. Some have suggested that PEL/VE are related pathophysiologically to

GM/IVH through an ischemic process [31, 32]. Other recent studies argue that
factors other than infarction or ischemia are important in the events leading to
white matter damage. The remote infection hypothesis suggests that fetal and
neonatal inflammatory responses damage developing white matter [33]. The
apparent association of NH and white matter damage may be explained by
several factors. Prematurity, infection, fetal growth, hypothyroxinemia, and
hypocarbia/hypercarbia may play roles in the relationship [34]. The prevalence of
PEL/VE is reported as 10%-15% in preterm babies. Neonatal intracranial
ultrasound abnormality, particularly white matter damage, is well documented to
be the most powerful predictor for development of CP [35, 36]. In follow-up
studies of this cohort, PEL/VE was found to predict a 15-fold excess risk of
disabling CP, and 5 times higher risk for non-disabling CP [29].

The time of onset of brain damage is a very critical point in any etiologic
study. Ultrasonographic diagnosis makes serial scanning over short intervals
feasible. Several investigators have performed timing and incidence studies of
neonatal brain damage. A substantial proportion of GM/IVH is seen on
ultrasounds obtained within 2-8 hours of birth, and most of the remainder has
developed by 3 days [37-39]. In the NBH cohort, we found that 34% to 44% of
GM/IVH, respectively, were present by first hour of life [40]. White matter
damage, however, usually can be detected by ultrasound at 7 days of life and
develops fully at several weeks of age. These data suggest that the period

around birth may be a critical time for the origin of neonatal brain lesions.

The effect of obstetric factors on fetus:

The reasons for excess risk of neurodevelopment impairment in preterm
and /or low birth weight babies have remained uncertain and controversial for
many years. As early as 1862, William John Little [41] recognized that medical
complications, especially perinatal asphyxia and birth injury, ensuing from the
preterm birth were the major factors leading to cerebral palsy. However, another
school of thought argued that neurological impairments likely stemmed from the
same events that lead to the preterm birth itself [42]. This thesis is carried out to
examine the roles of obstetric factors in adverse neonatal outcomes.

The passage through the birth canal is the first challenge for human
beings. The birthing experience has never been harm-free for mothers or babies.
The processes of labor and delivery is inherently hazardous. Obstetric practice
has developed dramatically in the past several decades with significant
decreases in perinatal mortality and morbidity. In recent years, interest has
focused increasingly on reexamination of the labor and delivery factors that may
play a part not only in survival of infants but also in the quality of life of survivors.
Fetal hazards of labor and delivery have received new attention, and physician’s
decisions about management of labor and delivery are being carefully
scrutinized.

The essential components of the course of labor are dilatation of the
cervix and fetal descent. Labor is traditionally divided into three stages. The first
stage is the duration from onset of labor to full cervical dilatation (100m). This
stage has been subdivided further into latent and active phases. The latent

phase is mainly as preparatory and involves little cervical dilatation. During the

active phase, cervical dilatation occurs at the most rapid rate. The second stage
of labor is from full cervical dilatation to birth of the baby. The third stage is from
birth of the infant to placental delivery.

The intensity of uterine contractions increases progressively throughout
labor [43]. Pathophysiological studies have found that each contraction impedes
of intervillous space blood flow. As progressive intramyometrial pressure
increases during contraction, the venous outflow is first obstructed, producing
congestion in the intervillous space and a decrease in matemal-fetal oxygen
transfer. Then the arterial inflow is also obstructed, so that the intervillous space
is physiologically isolated from the mother, and a transitory decrease of matemal-
fetal gas-exchange may occur. In addition, uterine contractions may be
associated with some degree of umbilical cord compression depending upon the
position of the cord [44]. The mechanical energy of the uterine contraction may
cause functional fetal “breath holding”. The labor process is thus a repetitive
mechanical stress to the fetus. Fetal blood gas studies show that fetal pH
decreases slowly during the first stage of labor and then falls more rapidly during
the second stage. A drop in fetal oxygen saturation of about 10% also occurs
during the second stage [45]. Continuous fetal heart monitoring shows that a
pattern of accelerations occurring with each uterine contraction seems to be the
earliest sign of fetal stress. This temporary reduction in oxygen transfer
associated with normal uterine contractions is generally of no clinical
significance. However, contractions superimposed on a fetus that is already

compromised (e.g. by prematurity, IUGR, infection), may jeopardized the

decreased margin of fetal reserve. The vulnerable fetus may not be able to
tolerate the stress of even a few minutes of labor. True fetal distress may occur.

Another formidable effect of labor is that it may result in tremendous
increases in cerebral venous pressure through deformations of the particularly
compliant premature skull [46]. These deformations are related to the magnitude
of intrauterine pressures, the duration of labor, and the fetal presentation. Such
deformations sometimes lead to obstruction of major venous sinuses, and
increase venous pressure [47]. The distinctive venous anatomy of the cerebrum
in preterm babies, thin walled veins of germinal matrix, is vulnerable to
hemorrhage at elevations of venous pressure. Observations of premature infants
have indicated that increased blood pressure from initially low levels may be
important in the pathogenesis of periventricular-intraventricular hemorrhage [48].
Animal studies in preterm sheep support the finding of frequent intracranial
hemorrhage when asphyxia and increased venous pressure occur together [49].
Based on these theories and observations, labor and delivery factors may be
associated with increased risk of neonatal brain damage and neonatal death as
well as later handicaps.

Low birth weight and /or preterm birth occur from three clinical situations:
preterm labor, preterm premature rupture of membranes (PROM), and physician-
initiated delivery for matemanetal complications. Preterm labor and PROM
frequently associated. Labor ensues in approximately 80% of PROM patients
within 48 hours after rupture of membranes. A small proportion of PROM leads to
preterm delivery through iatrogenic intervention because of concern about

subsequent infection. The trigger for preterm labor is not clear. Multiple factors

are thought to be involved in the initiation of human preterm labor. Prostaglandin,
catecholamines, increased uterine activity, premature cervical ripening,
alterations in estrogen-progesterone ratios, and changes in uterine blood flow
have all been implicated in the initiation of preterm labor [50]. Maternal
complications, such as abruptio placenta, uterine anomaly, multiple gestations,
and amnionitis are also associated with preterm labor[51, 52]. Increased uterine
activity, beginning in the second trimester of pregnancy, long prior to overt
preterm labor has been noted in women with preterm labor. Thus, a fetus
experiencing preterm labor is either in an unfavorable uterine environment or has
been exposed to a prolonged period of uterine contraction. These factors may be

associated with adverse neonatal outcomes.

Previous studies:

Controversy persists regarding the roles of presence of labor and route of
delivery in the occurrence of neonatal brain damage and cerebral palsy (CP) in
the low birth weight population. Much of this controversy stems from the difficulty
of controlling for multiple confounders and the difficulty of conducting a clinical
trial to assess this problem. Many studies suggest that low birth weight infants
born after labor are at increased risk for GM/IVH. Other studies report that there
is no such relationship between labor and brain hemorrhage. The relationship

between obstetric factors and white matter damage is even less clear.

Table 1.1: Previous studies:

 

 

 

 

 

 

 

 

cozmoiEoo .323: .m 38-8
.Emou .9382 .m H.4.me So
.908 Emma< .F m\o .m> 359.5 83022,:
Econ .3382 .N m 8348 Em.
.0: 3m :2; .259 .9 m6 8: Ba;
5wa 38:82 .N
.ommEoEo: ma .m m ooomuv
BEoEcggE 6888 So. .N 5.3
sz mcozL 1930:2923. .Fmaomcmanm 85.02 .F 0mm ENE...
6.89.2 .mc_mm>
Sooodvav 6822 o: .m:_mm>
88.2---593 m>zom comm. m>zom m5
m5: .mc_mm> ionm. comm. EBB. m5
EBB. m5 loom. o; .m> Loam. oz.m
£9; 2 26: =52:ch C: .E .m> Long 5:5 m5
moofla .m mwooxm $3 a . H v 12 963.505 .mdionm. 59:? Q0 .w m omtuv Fméfl
wz .N .2583 can k .12 em. .m .82 9:8 .965 a
ooofia .F .>>m .xmm 5 TV 12>.me— .F o: @3092 m>=o< .— omm c8893
>_mmmEo_:oEcm>.v
.2 720mm >9 85qu 20mm .m_om_mon:m_ carbon”:
anEO .N .m BEoEcmea .m ucm nmwuvLonfl moomToom
oéumO .FEoLBmooEoo .Emcmnozoo .N .w> Loam. 02 .N Tea
H__mcm>O .<O .>>m .I>_.F m5 .m> .mc_mm> .P 39 :3ch
.82:
52.20 as... .85 6 a 88v Ea
Amdé. 5 mNumo <0 .26 5:83 new 85me m3. c233
gamma EcEfiig mEooSO mSmoaxm $835+ 853‘

 

 

 

 

 

 

 

+ sample size and birthweight/gestational age

* randomized trial.

 

 

 

 

 

 

 

 

Table 1.1: Previous studies: (cont’d)

 

 

 

 

 

b__m...oE Emcee: Eczflcomoa :398 Egg,
.8. v n. 952. o...88-.8_m§....m m6 .2, .a.._8> 5.5 __< .8. 8..
.mz .m
.58. v a. ...>_ .N 32.3883 88.... a 83-08 .8.
.88 .8 ask. K .. 802. 58.. .238: .. m5 .8 88> 8m. 52.
:8... v n. :3
.0840. a, «8.2 a 1228:8898...
08.. a, .8. z .m .V v.82 558 .m
.8... .9, ask. a .m .V v.82 .._E F .m 32.988... 885. o 89v
OER .9 08.3 .. 802 £89.. .9882 4 m5 .8 .m.._8> 82.8. 8.8
m5
.3 mcozmofiE
.om.o-mm.o. mm... 98 8.6.88...
”comes c. gonm. m mmiéom
83.8.9 8.. 3.89. at...
”85> c. .o .99. .>>m .<0 b..mtoE .9382 m5 .m> .mc_mm> monm $8 $5..
m_._co_cEmo_.oco
uoEmEnoou £032.33. 2. mm-mm
€3-38. wwoumo 5.; 8.32.2828 m6 .8 88> 8 .8. Sam
.888 .N .08. o... .9588 a 08.88
.Soodvh. .. .802 :>_ .8829 mo. mom .8. .85.
4.5.5.839
.o.oEoco>mn. .Em. Emu 5.5 .8sz32
.0. nova. .Nucm .938; .<o .o.oEo_.o>wn..N m ooomv .5.
mo .0 a... oz .. .Bm 5. 8.22. no. ma .8 .859 68> 8 8.982...
«.zmom EmEfiPfia. mEooSO Samoaxm. $8.339. .oE:<

 

+ sample size and birthweight/gestational age

* exposure control study.

 

 

10

Leviton et al’s [53] retrospective study found that VLBW babies delivered
vaginally were more likely to have germinal matrix hemorrhage, and the risk
increased with increasing duration of labor. No multiple regression was done to
control the confounders in their study. However, Hanson et als [54] prospective
cohort study of 500-1 ,SOOg birthweight infants addressed the relationship
between labor and delivery characteristics (duration of labor, interval between
membrane rupture and delivery, and route of delivery) and cranial
ultrasonographic abnormalities (GM/IVH, moderate to severe ventricular
enlargement, echodensity, and periventricular leukomalacia). They found that
neither presence of labor, nor duration of labor, was related to brain damage.
Vaginal delivery was the only obstetric characteristic consistently associated with
increased intracranial hemorrhage and white matter damage. This relationship
was, however, markedly reduced when placental inflammation was accounted
for. They argued that vaginal delivery might have acted as a marker of infection.
However, Anderson and Shaver et al [55-57] reported that active labor was
related to increased risk of early onset brain hemorrhage and progressive (grade

IIIor IV) hemorrhage compared to latent labor and no labor. This risk was not

seen for late onset hemorrhage. Their study also showed risks of early GM/IVH
dependent on delivery mode and labor, with cesarean section without labor
representing the lowest risk, and cesarean section with active labor, and vaginal
delivery without forceps representing the highest risks.

Several studies [58-60] concerning delivery mode in LBW infants
suggested that cesarean delivery is not associated with a lower risk of either

mortality or neurodevelopmental abnormalities. However, in a exposure control

ll

study, a higher rate of psychomotor retardation [61] was related to vaginal
delivery with vertex presentation compared to cesarean section delivery. Other
studies indicated cesarean section did have a protective effect on brain
hemorrhage with statistical significance [62, 63]. Studies focused on breech
presentation found that cesarean birth is associated with improved survival in
VLBW infants. No benefit of reduced IVH or CF was seen in these cesarean
section delivered breech presentations [64-67].

Most studies to date have examined labor and delivery effects on brain
injury without consideration of possible confounding factors, which may also
contribute to poor outcomes. Although birthweight and gestational age are
commonly considered for adjustment in most studies, authors have ignored other
important potential confounders, which may substantially affect the relationship of
obstetric factors and neurodevelopment outcomes. For example, few studies
have considered the effect of pre-eclampsia in analysis. Pre-eclampsia have
often been found to be associated with a reduced risk of brain lesions [68, 69]
and OF [70-72] in low birthweight babies. Pre-eclampsia is one of the
predominant indications for preterm birth, and delivery is often therapeutically
decided by physicians. Pre-eclampsia is therefore related to less labor and to
more use of an operative delivery mode. Maternal infection is also thought to be
related to neonatal outcomes. Amnionitis is a predictor of neonatal brain damage
and CP. It is also associated with two leading causes of preterm birth: preterm
labor and premature rupture of membranes. Placenta abruptio can lead to
preterm labor and usually results in delivery by cesarean section. It has also

been found to be associated with CP [73]. Failure to control for such confounding

12

factors may produce false results. In addition, the sample size was too small in

some studies to reliably detect effects of labor and delivery.

CHAPTER 2

MATERIALS AND METHODS

:3de Popglation:

The Neonatal Brain Hemorrhage Study (NBH)[24] is a longitudinal cohort
study aimed at assessing the causes and outcomes of germinal matrix/
intraventricular hemorrhage and other brain lesions in low birth weight infants.
Study subjects were recruited from three hospitals with neonatal intensive care
units (NICU) in central New Jersey: Jersey Shore Medical Center, Neptune
(JSMC); Monmouth Medical Center, Long Branch (MMC); and St Peter’s Medical
Center, New Brunswick (SPMC). Between Aug 27, 1984 and Jun 30, 1987, all
infants weighing 501 - 2000g at birth admitted to these three NICUs were eligible
for enrollment. There were 55,107 babies born during that period in the study
area. Of these, 1,318 (2.4%) were born weighing between 501 and 2,0009 and
1,105 infants were enrolled in NBH. Thus, 84% of all the eligible children in the
design were enrolled in NBH. Of these 1,105 infants, 982(89%) were born in, and
123(11%) were transferred to, the three study hospitals. Mean birthweight was
1,3939 (i4069) and the mean gestational age was 30.9weeks (:36 weeks).
Twenty six percent of the enrollees were black, and 25% were products of

multiple pregnancies.

l4

Outcome assessment:

Neonatal brain lesions, neonatal death and CP are selected health
outcomes. Neonatal death was defined as death no later than 28 days after birth.
Neonatal brain lesions were diagnosed from ultrasonographic screening following
a predesigned protocol, and CP was assessed at 2 years of age.

Each infant was prospectively screened by ultrasound examination for the
presence of brain lesions. The ultrasound protocol initially called for three
postnatal cranial ultrasound scans, timed as closely as possible to 4 hours, 24
hours and 7 days after birth. Among 1,105 enrolled infants, 1,088 (98.5%)
received at least one protocol ultrasound scan; three infants had only a single
non-protocol scan. The remaining 14 infants died before the first scan could be
obtained. One or two further scans were later added to be obtained between 11
days and discharge. 517 infants (46.8%) received these later scans to capture
lesions not imaged in the first week of life. The pool of readers included five
hospital radiologists and two outside experts. Each scan was read by two
independent readers. Discordant readings were submitted to a third reader for
arbitration. Readers were blind to the infant’s clinical condition and to the other
reader’s interpretation. Ultrasonographic diagnoses were categorized as:

1. Gerrninal matrix/intraventricular hemorrhage (GM/IVH).
These diagnoses were made when a focal area of increased echogenicity in the
thalamocaudate groove or extending to the head of the caudate nucleus and / or
an echogenic focus or foci within the lateral, third or fourth ventricles separate
from the choroids plexus were seen on the ultrasonographic films. lVH was also

diagnosed when irregularity of the choroids plexus margin indicated adherent

15

intra-ventricular blood. This category corresponds with hemorrhage grades I
and II, respectively, in the Papile classification [74].

2. Parenchymal echodensities/lucencies and /or ventricular enlargement
(PEL/VE).
This category was defined as abnormal parenchymal echogenicity seen in the
hemispherical white matter and / or any enlargement of the third, fourth, or each
lateral ventricle. It includes periventricular leukomalacia, porencephalic cyst, and

grades III and IV IVH.

In 1019 subjects, ultrasound diagnosis is based on the consensus of two
readers. In this analysis, the consensus diagnosis is used. 262 GM/IVH and 112
PEL/VE were diagnosed in the neonatal period in these subjects. Only 14% of all
the PEL/VE had first detected on the later scans [75].

At two years follow-up, 901 children had survived and 777(86%) of them
were successfully contacted. 80% of survivors were examined by a pediatric
nurse or a pediatric nurse practitioner, specially trained for the project. Another
6% of survivors were contacted by phone interview and medical record review.
The diagnosis of CP was based on review of specific neurologic assessment by
the pediatric nurse/pediatric nurse practitioner, including semi-quantitative
assessment of tone and of extrapyramidal movements; tendon reflexes; shoulder
flexibility; primitive reflexes; cranial nerve abnormalities; goniometric
measurements of knee, ankle, and hip ranges of motion. Children who were
diagnosed as having CP were further classified as having either disabling CP

(DCP) or non-disabling CP based on the following criteria . If the child had at

16

least one of five markers of disability in addition to specific neurologic findings,
the diagnosis of DCP was given.

1. Inability to walk ten steps unaided by 2 years of age;

2. Bayley motor score greater than 1 SD lower than performance score;

3. Receipt of physical therapy for motor disability;

4. Receipt of surgical intervention for motor disorder;

5. Use of braces or physical assistance devices.
Bayley testing was performed by a licensed clinical psychologist. Motor status
was quantitatively measured using the Bayley Psychomotor Development Index

(PDI), which is a standardized test with a mean of 100 and SD of 16 [76].

aposure assesiment:

The exposures of interest in this thesis are labor & delivery status and
initiation of preterm labor. In NBH cohort, medical information for study subjects
was compiled daily from medical charts by seven trained nurses. Validation of
data collection uniformity and accuracy was performed periodically. Delivery
mode was recorded as vaginal or cesarean section. Vaginal delivery was
specified as spontaneous, forceps or vacuum extraction. Episiotomy and
lacerations during delivery were recorded. If infants were delivered by cesarean
section, indications for cesarean section were noted. Following tradition, labor
was classified in three stages, and the first stage was further classified as latent
phase and active phase. Latent phase was defined as beginning with “onset of
regular contraction”, and active phase was defined as beginning with “cervical

dilatation >3 cm, effacement >80%, contractions every 3 mins, for at least 1 hour.

17

All conditions must be present.” Labor status was noted at transport or hospital

admission. In labor & delivery records, details of onset date, time and duration of

each phase and stage of labor were abstracted. Cervical dilatation, effacement

and examination time were recorded on admission, mid point and prior to

delivery. If oxytocin was used, the indication, start time and stop time, dosage

and cervical status before oxytocin admission were also noted. Table 2.1 lists the

criteria and numbers of subjects classified according to each definition.

Table 2.1: Criteria for the classifications of labor and delivery variable:
1. No labor:
1. Delivered by cesarean section.
AND
2. No evidence of any labor.

2. C/S with latent labor:
1. Delivered by cesarean section.
AND
2. Had at least one of the following signs of latent labor.
a. Recorded start time and/or duration of latent labor.
b. Either used tocolysis or used oxytocin (but not for OCT).
c. Indication for cesarean section was “failure to progress”.
d. Primiparous women who had cervical dilatation no more than
3 cm.
9. Progressive cervical dilatation from 1cm to 3cm between two
examinations.
f. Second twin, when the first twin had latent labor.
AND
3. No evidence of active labor.

3. Active labor with cesarean section:
1. Delivered by cesarean section.
AND
2. Had at least one of the following signs of active labor.
a. Start time recorded either for active phase or second stage.
b. Cervical dilatation at least 4cm.
c. Second twin, when the first twin had active labor.

4. Vaginal delivery.

5. Unable to classify due to insufficient information.

182

199

121
68

199

125
56
18
478

47

The classification of the labor & delivery variables is fairly straightforward.
Delivery methods are simply classified as cesarean section or vaginal delivery.
Cesarean section patients are further classified into 3 groups according to labor
status based on the traditional classification of labor. The main resource for this
classification was records during the labor course. If these records were not
available, cervical dilatation, oxytocin or tocolysis admission, and indications for
cesarean section were used to determine the labor status.

The process of initiation of preterm labor was also assessed in the infants
who experienced labor. The effects of iatrogenic preterm labor are compared the
effect of spontaneous preterm labor (either idiopathic or caused by matemal/fetal
complications). This classification is based on information on indications for
oxytocin use and labor status (Table 2.2). If labor was induced by oxytocin, it is
classified as iatrogenic. Otherwise, the labor is considered spontaneous.

Table 2.2: Criteria for the classification of cause for preterm labor:
Among subjects with labor.

 

1. Spontaneous preterm labor: 817
Any one of the following: '
a. Tocolysis was used (including Mg+ and no hypertension). 292
b. Admitted to hospital because of preterm labor. 282
0. Never used oxytocin for induction of labor. 227
d. Oxytocin was used after labor onset. 12
e. The reason for hospitalization was PROM. 4
2. Iatrogenic preterm labor: 59
Any one of the following:
a. Oxytocin used for induction because of matemal/fetal 49
complications.
b. Oxytocin used before labor onset and not for OCT. 3
c. Admission to hospital because of hypertension or IUGR 7

and no labor present upon admission.

3. Unable to classify due to insufficient information. 47

 

l9

Other birth characteristics:

Birth weight was measured after delivery. Gestational age was established
by a hierarchical algorithm based on several data sources (prenatal ultrasound
scans, prenatal medical records, postnatal maternal interview, and labor &
delivery records) [77]. Fetal growth was measured as fetal growth ratio (FGR),
which was defined as the ratio of birth weight to the median birth weight for
completed gestational weeks, using the standards developed by Yudkin et al
[78]. Infants were categorized as small for gestational age (SGA) if the fetal
growth ratios were below the 10th percentile [75]. Maternal hypertension was
classified into preexisting hypertension, pregnancy-induced hypertension and
pre-eclampsia based on the combination of prenatal and intrapartum blood
pressure measurements, clinical diagnosis and preexisting history [79].
Matemanetal complications, including IUGR, PROM, amnionitis,
placenta/umbilical cord abnormality, and fetal presentation were abstracted from
medical records. These variables are purely from clinical diagnosis in medical
charts. Maternal interview provided information on race, age, parity, education,
smoking, previous pregnancy outcomes and pre-pregnancy history. The neonatal

variables were abstracted from initial resuscitation records and ICU chart.

Statistical analysis

Because many of the obstetric items were categorical, group comparisons

were carried out by simple cross tabulations and f test of proportions. One-way

analysis of variance with multiple comparisons was performed to evaluate the

means and standard deviations of numerical variables among groups.

20

Unadjusted risks of outcomes in stratified analysis were reported as percentages.
Both univariate and multivariate logistic regression models were developed.
Those variables associated with both the outcomes and the exposures either in
our data or from other literature were considered potential confounders and were
entered into the multiple logistic regression along with interactions between the
exposure and the potential confounders. The variables that remained in the final
model were the labor and delivery method , birth weight, gestational age, multiple
gestation, maternal hypertension, amnionitis and placenta abruptio. Separate
models were run for infants with and without SGA. All data analyses were carried

out using the SAS statistical package (SAS Institute, Cary, NC) version 8.

21

CHAPTER 3

RESULTS

Sample characteristics:

Of the 1,105 infants enrolled in the study, 1,058 subjects had detailed
intrapartum information. Of the 47(4%) with missing information, 34 babies were
transferred from other hospitals. The following analyses were based on 1,058
patients (Figure 3.1).

Prior to age 2 follow up, 130(12.3%) neonatal deaths (<=28 days) and an
additional 48 deaths occurred. GM/IVH was found in 245(23.2%) neonates and
PEL/VE developed in 101 (9.6%) neonates. Either missing consensus ultrasound
diagnosis or death before first ultrasound schedule occurred in 81 babies.
111(10.5%) children were diagnosed as having CP at age 2 and 60(5.7%) of
them were classified as having disabling CP. 175 children were lost to follow up
for neurological outcomes at CP assessment.

The mean birthweight was 1,401i407g, and the mean gestational age
was 31i4 weeks. The overall vaginal delivery rate was 45.2%(478) and cesarean
section was 54.8%(580). Thirty eight percent of infants were delivered by
cesarean section with labor, half of whom (18.8%) were delivered with latent
labor and half (18.1%) with active labor. The remaining 17.2% were born by
cesarean section without labor. In patients with labor, 817 preterm births
experienced spontaneous preterm labor, and, for the remaining 59, the labor was

iatrogenic.

22

Figure 3.1

Distributions of exposure and outcomes

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Total Sample
1,105
Missing: 47
GM/IVH
245(23.2%)
PEUVE
101 (95%) Neonatal deaths
0
. . Sample for 130(12.3 /o)
Missmg: 81 _, analysis: __
1,058
Later deaths
48
CP
1 1 1 (10.5%)
DCP
60(5.7%)
Missing: 175
No labor C/S + latent labor C/S + active labor Vaginal
182 (17.2%) 199 (18.8%) 199 (18.1%) 478 (45.2%)
Spontaneous Iatrogenic
preterm labor preterm labor
817 59

 

 

 

 

 

 

These outcomes (deaths, GM/IVH, PEUVE and CP) are overlapped.

23

Table 3.1: Characteristics of cohort:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cesarean Section Vaginal
No Lab Latent Active Delivery
*PTD: Iatrogenic - 15% (30) 2% (4) 5% (25)
Spontaneous 85% (169) 98% (195) 95% (453)
Birth weight 1420:367 1450:365 1402:372 1374:450
*Best GA 32:3 32:3 31 :3 30:4
Sex Boy 47% (85) 51% (102) 54% (107) 53% (252)
Girl 53% Q7) 49% (91) 46% (92) 47% (226)
Race White 72% (131 ) 74% (147) 76% (151 ) 72% (344)
Black 28% (51) 26% (52) 24% (48) 28% (134)
*Multiple Yes 23% (42) 36% (72) 52% (103) 13% (61)
No 77% (140) 64% (127) 48% (96) 87% (417)
*Pre- No hyper 41% (74) 74% (147) 82% (164) 87% (414)
eclampsia Preexist hyper 11% (20) 3% (5) 3% (6) 3% (15)
PIH 15% (28) 7% (14) 6% (11) 5% (25)
Pre-eclampsia 33% (60) 17% (33) 9% (18) 5% (24)
*PROM >24hours 23% (33) 17% (30) 21 % (37) 31% (140)
Amnionitis 5% (8) 6% (11) 5% (10) 5% (23)
*Vertex 48% (59) 52% (79) 45% (75) 87% (355)
present
*Oxytocin 4% (7) 19% (37) 9% (18) 24% (113)
*lndications Fetal distress 25% (40) 27% (48) 21% (36) -
for Progress 0 5% (9) 5% (9)
cesarean Placenta 14% (22) 15% (27) 5% (9)
section Malpresent 18% (29) 28% (49) 46% (80)
Other 43% (70) 25% (45) 23% (40)
*SGA 44% (80) 45% (90) 28% (56) 24% (117)
*Placenta 6% (11) 7% (14) 4% (8) 1% (2)
previa
*Placenta 1 1% (19) 16% (32) 6% (11) 6% (28)
abruptio
Meconium 8% (14) 5% (10) 5% (10) 6% (28)
Apgar1 6:3 6:3 5:2 6:3
Apgar5 8:1 8:2 7:2 7:2
Vent days 6:1 8 7:20 1 0:20 8:21
Oxygen 10:23 11:25 15:29 11:26
days
CLD 8% (14) 9% (16) 16% EL 12% (49)

 

*There are statistically significant differences among four groups (p<0.05).

24

 

Table 3.1 shows the population categorized into different exposure
classifications. Among these exposure classifications, there was no significant
difference regarding maternal education, pre-pregnancy characteristics and prior
pregnancy history (data not shown in table). Gender and race were equally
distributed in each stratum of the classifications. Active labor was most likely to
be of spontaneously onset. Iatrogenic labor was most likely to result in C/S
delivery. Babies delivered by cesarean section were slightly heavier and more
mature than those delivered vaginally, but the birthweight differences did not
reach statistical significance. The likelihood of C/S was significantly higher in
multiple gestations than in singleton gestations, especially after labor onset.

Among pregnancy complications, maternal hypertension was significantly
associated with labor and delivery factors. Delivery by US without labor was
most likely with preexisting hypertension and pregnancy-induced hypertension/
pre-eclampsia. Infants with prolonged PROM and vertex fetal presentations were
most likely to be delivered vaginally. There was no significant difference in
PROM and fetal presentation among cesarean section groups. The percentages
with amnionitis were similar in each exposure group. The most frequent
indication for cesarean section without labor was “other" category, including
maternal/fetal complications. Latent labor ending cesarean section was most
commonly due to fetal distress or abnormal fetal presentation. Almost half the
unsuccessful active labors were due to abnormal fetal presentation. Nearly half
of the infants deliver by C/S without labor or after latent labor were small for
gestational age infants compared to approximately a quarter of infants delivered

after active labor. Placenta previa or abruptio was more common with cesarean

25

section after no labor or latent labor than they were with cesarean section after
active labor or with vaginal delivery.

There was not much difference in Apgar scores, duration of ventilation
and oxygen use in the neonatal period among these classifications. However,
infants that experienced active labor had a higher rate of chronic lung diseases

(CLD), but not statistically significant.

Risks of outcomes:

Tables 3235 list the percentages of specific outcomes within stratified
cells. Younger gestational age, lower birth weight, abruptio placenta, lower Apgar
scores and chronic lung disease were consistently associated with increased
risks of all neurodevelopment abnormalities and neonatal death. Amnionitis was
related to a higher risk of GM/IVH and DCP, most dramatically in patients
delivered by cesarean section with active labor. However, amnionitis was not
related to PEL/VE and neonatal death. Pre-eclampsia was related to decreased
risk of brain damage (both GM/IVH and PEUVE). However, this benefit was not
found in cesarean section with active labor. Pre-eclampsia also showed a
protective effect for DCP. SGA babies tended to have lower risks of brain
damage and DCP. This relationship did not appear in patients with active labor
cesarean section, when brain damage was the outcome. Neonatal brain damage

was associated with more deaths in the cohort.

26

Table 3.2: Unadjusted cell specific risk (%) of GM/IVH.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cesarean Section Vaginal
No Lab Latent Active Delivery
Birth weight <1000 27 33 44 50
1 000-1499 20 27 43 33
>=1500 10 11 18 13
Best GA <28 33 27 36 54
28-<35 16 22 32 19
>=35 9 3 26 6
Sex Boy 14 24 33 30
Girl 19 15 31 25
Race White 13 17 31 26
Black 27 28 34 32
Multiple Yes 1 1 25 36 16
No 18 17 27 29
Pre-eclampsia No hyper 23 22 35 29
Preexist 21 0 17 21
PIH 22 25 18 29
Preeclamp 5 9 17 10
PROM Yes 26 17 36 33
>24hours No 15 22 31 26
Indications for Fetal distress 16 19 29 -
C/S Progress 0 22 25
Placenta 19 27 22
Malpresent 21 16 38
Other 15 26 28
Amnionitis Yes 38 27 67 48
No 15 19 30 27
SGA Yes 12 14 30 17
No 21 25 33 31
Oxytocin Yes 29 1 1 13 25
No 16 22 34 28
Presentation Vertex 20 21 38 26
No vertex 21 21 30 49
Placenta Yes 1 1 29 13 50
previa No 17 19 33 28
Placenta Yes 22 20 45 35
abruptio No 15 20 31 28
Apgar1 <7 21 29 41 40
>=7 12 10 15 17
Apgar5 <7 35 36 39 46
>=7 14 16 30 21
CLD Yes 21 36 54 50
No 16 18 ' 28 25
Died Yes 21 24 47 54
No 16 19 29 21

 

 

 

 

 

27

 

Table 3.3: Unadjusted cell specific risk (%) of PELVE.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cesarean Section Vaginal
No Lab Latent Active Delivery
Birth weight <1000 8 11 24 21
1000-1499 6 12 18 15
>=1500 5 5 8 4
Best GA <28 0 18 24 23
28-<35 7 8 13 7
>=35 5 3 13 6
Sex Boy 8 11 16 13
Girl 4 7 13 9
Race White 6 7 14 10
Black 6 12 15 14
Multiple Yes 3 9 13 14
No 7 8 15 11
Pre-eclampsia No hyper 9 10 15 12
Preexist 16 0 17 7
PIH 0 8 9 10
Preeclamp 2 3 11 0
PROM Yes 10 0 6 13
>24hours No 3 12 17 1 1
Indications for Fetal distress 5 11 15 -
C/S Progress - 0 13
Placenta 14 19 11
Malpresent 0 7 14
Other 3 7 13
Amnionitis Yes 0 0 22 24
No 5 9 14 11
Oxytocin Yes 14 3 13 4
No 6 10 15 13
SGA Yes 3 7 16 8
No 9 10 14 12
Presentation Vertex 5 9 15 9
No vertex 3 9 13 20
Placenta Yes 0 21 1 3 0
previa No 5 8 15 11
Placenta Yes 17 17 27 26
abruptio No 4 7 14 11
Apgar1 <7 8 14 20 17
>=7 4 3 4 6
Apgar5 <7 13 18 30 20
>=7 5 6 10 8
CLD Yes 21 14 28 23
No 5 6 9 7
Died Yes 7 24 34 25
No 6 7 10 8

 

 

 

 

 

28

 

Table 3.4: Unadjusted cell specific risk (%) of Neonatal death.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cesarean Section Vaginal
No Lab Latent Active Delivery
Birth weight <1000 22 12 32 44
1000-1499 7 6 10 5
>=1500 1 0 9 7
Best GA <28 25 27 36 54
28-<35 6 22 32 19
>=35 0 3 26 6
Sex Boy 4 7 11 18
Girl 9 5 15 15
Race White 7 7 13 16
Black 6 4 13 19
Multiple Yes 7 8 13 23
No 6 5 14 16
Pre-eclampsia No hyper 8 8 13 18
Preexist 10 0 0 13
PIH 4 0 18 12
Preeclafl) 5 0 11 4
PROM Yes 9 3 11 14
>24hours No 5 7 10 18
Indications for Fetal distress 8 8 25 -
C/S Progress 0 0 0
Placenta 5 7 11
Malpresent 7 4 9
Other 7 2 15
Amnionitis Yes 0 0 20 22
No 7 6 13 17
Oxytocin Yes 14 0 6 14
No 6 7 14 18
SGA Yes 4 8 16 15
No 9 5 12 17
Presentation Vertex 3 8 17 13
No vertex 9 4 5 45
Placenta Yes 9 0 0 50
previa No 7 6 14 17
Placenta Yes 5 13 27 21
abruptio No 7 5 12 17
Apgar1 <7 13 10 18 29
>=7 0 2 4 5
Apgar5 <7 22 14 33 42
>=7 4 4 7 8

 

 

 

 

 

29

 

Table 3.5: Unadjusted cell specific risk (%) of DCP.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cesarean Section Vaginal
No Lab Latent Active Delivery
Birth weight <1000 6 22 33 14
1000-1499 5 7 13 15
>=1500 5 4 6 2
Best GA <28 1 1 27 18 20
28-<35 5 4 12 6
>=35 5 7 6 0
Sex Boy 8 7 12 5
Girl 4 7 12 10
Race White 8 7 12 8
Black 0 8 9 5
Multiple Yes 0 8 14 8
No 7 6 9 8
Pre-eclampsia No hyper 10 10 14 8
Preexist 7 0 20 9
PIH 0 0 0 5
Preeclamp 2 0 0 0
PROM Yes 4 5 11 3
>24hours No 6 7 11 9
Indications for Fetal distress 10 8 11 -
C/S Progress - 0 0
Placenta 18 17 0
Malpresent 0 6 14
Other 0 3 3
Amnionitis Yes 20 13 50 6
No 4 7 8 7
Oxytocin Yes 0 3 7 5
No 6 8 12 8
SGA Yes 2 5 5 5
No 9 8 14 9
Presentation Vertex 4 8 5 6
No vertex 4 5 15 11
Placenta Yes 0 25 0 0
revia No 5 6 11 7
Placenta Yes 29 19 14 17
abruptio No 2 5 10 7
Apgar1 <7 7 10 15 14
>=7 4 4 7 3
Apgar5 <7 0 19 21 11
>=7 6 5 10 7
CLD Yes 40 42 47 28
No 3 4 6 5

 

 

 

 

 

3O

 

Active labor, regardless of delivery mode, was consistently associated
with elevated risks of adverse neonatal outcomes compared to no labor or latent
labor only. For neurodevelopment outcomes, cesarean section with active labor
was related to the highest risk in most strata, especially when combined with
other risk factors (e.g. amnionitis, placenta abruptio, CLD). Vaginal delivery was
associated with more neonatal death than was cesarean section delivery. These
patterns were not as clear for DCP.

The effect of cesarean section with active labor on neonatal outcomes
was clearly stratified by SGA. SGA infants with active labor were more vulnerable
to developing adverse outcomes compared to no labor (not DCP). In children
who died before 2 years follow-up, relationships between exposures and PEL/VE
were much stronger than in surviving children (Table 3.3).

Table 3.6 shows the percentages of the four outcomes of interest in each
labor group (excluding the no labor group), comparing infants whose labor was
spontaneous with all labors. Cesarean section with active labor had the highest
rates of GM/IVH and PEUVE. Cesarean section with active labor and vaginal
delivery were both associated with more neonatal deaths. No difference in risk of
DCP was seen among these groups. When this analysis was restricted to infants
who had spontaneous labor, the relationships of labor & delivery status and

outcomes did not change.

31

Table 3.6 Adverse outcomes in relation with labor & delivery comparing by any
labor and spontaneous labor only.

 

 

C/S with latent C/S with active Vaginal delivery
labor labor

With any labor
GM/IVH 20% (37) 32% (60) 28% (120)
PEUVE 9% (16) 14% (27) 1 1% (48)
Neonatal death 5% (10) 12% (22) 14% (61)
DCP 6% (11) 7% (13) 5% (22)
With spontaneous
labor only
IVH 21 % (34) 32% (59) 28% (116)
PELVE 9% (15) 14% (25) 12% (48)
Neonatal death 6% (1 0) 12% (22) 14% (57)
DCP 6% Q0) 7% (13) 5% (22)

 

 

Univariate and Multivariate analvsis:

Tables 3.7-3.10 list the odds ratios and 95% confidence intervals from
adjusted and unadjusted logistic regression models for the four outcomes. Model
1 divided the labor and delivery exposure into four categories. Because of the
similar characteristics of the cesarean section with active labor and vaginal
delivery groups on the one hand, and in the no labor and latent labor only
groups, on the other hand, the first two and the last two categories were
combined to create 2 divisions, active labor versus no active labor, in model 2.

Table 3.7 illustrates the analysis of risks for GM/IVH. Active labor,
especially were accompanied by cesarean section, represented significantly
increased risks for GM/IVH (OR=1.9-2.3). After adjusting for birthweight,
gestational age, multiple gestations, and other obstetric variables (maternal

hypertension, amnionitis and placenta abruptio), the odds ratio of cesarean

32

section with active labor decreased but remained significant. The significance of
vaginal delivery, however, disappeared. Pre-eclampsia played a protective role,
and amnionitis accounted for a significantly increased risk for GM/IVH. After
stratifying the models by SGA, the ORs of active labor didn’t change much
although the significance reduced. Only amnionitis had a significant effect in AGA
babies. There was no modified effect of SGA in the relationship between labor
and GM/IVH. When the SGA variable was added in the multivariate models for
the whole cohort, the ORs of cesarean section with active labor (in model 1) and
active labor (in model 2) did not change and remained significant (data not
shown in table).

Table 3.8 shows that cesarean section with active labor had a higher risk
for PEL/VE (OR=1.7-2.6) compared to no labor and this relationship remained
significant after controlling for potential confounders. Placenta abruptio also had
significant risk for PEL/VE, but amnionitis did not. These relationships were
modified by SGA status. The effect of cesarean section with active labor (OR
increases from 2.6 to 5.4) doubled in SGA babies and disappeared in AGA
babies. The effect of active labor versus no active labor was also increased in
SGA babies and remained borderline significant. OR for amnionitis (OR
increases from 1.3 to 6.5) on PEUVE increased dramatically and became
significant in SGA. Placenta abruption was only significant in AGA infants. The
95% confidence intervals were wider in stratified analysis because of the small
SGA sample size.

Active labor, regardless of delivery mode, showed twice the risk for

neonatal death than no labor (table 3.9). After controlling for confounders

33

(birthweight, gestational age, multiple gestation, maternal hypertension,
amnionitis and placenta abruptio), no factor remained significant. In SGA babies,
both cesarean section with active labor and vaginal delivery were strong risks for
neonatal death (OR=4.4). Active labor showed twice the risk for neonatal death
compared to no active labor in both uncontrolled and controlled models. This
relationship only remained significant in SGA babies.

No labor factors were related to disabling CP (table 3.10). Amnionitis and
placenta abruptio carried 4-5 times the risk for DCP, and pre-eclampsia showed
a protective effect for DCP. Univariate model was stratified by SGA. The ORs in
SGA were much larger than that in no SGA, although they were not statistically
significant. There was no significant effect of labor on disabling CP in subgroups
with different gestational age (data not shown).

In logistic regression models, spontaneous preterm labor was not related
to any neurodevelopment outcomes or neonatal death in either adjusted or

unadjusted analysis (data not shown).

34

Table 3.7 Logistic regressions for GM/IVH:

 

 

 

 

Model 1:
Univariate model Multivariate models

OR 95% Cl OR 95% Cl
C/S with latent labor 1.23 0.72-2.12 1.12 0.62-2.03
C/S with active labor 2.34 1.41 -3.90 1.96 1.10-3.49
Vaginal delivery 1.91 1.21 -3.02 1.37 0.81 -2.32
Pro-eclampsia 0.36 0.18-0.72
Amnionitis 2.32 1 .21-4.44
Placenta abruptio 1.27 0.72-2.23

Stratified by SGA
SGA AGA

OR 95% Cl OR 95% Cl
C/S with latent labor 1.04 0.37-2.89 1.09 0.51-2.34
C/S with active labor 2.29 0.84-6.21 1.67 0.81-3.44
Vaginal delivery 1.71 0.66-4.40 1.16 0.60-2.26
Pre-eclampsia 0.46 0.17-1.19 0.40 0.14-1.12
Amnionitis 3.78 092-1556 2.12 1.01 -4.42
Placenta abruptio 1.84 0.71-4.79 1.05 0.51 -2.15
Model 2:

Univariate model Multivariate models

OR 95% Cl OR 95% Cl
Active labor vs. no 1 .82 1 32250 1.47 1 02-213
active labor
Pre-eclampsia 0.37 0.19-0.73
Amnionitis 2.37 1.24-4.54
Placenta abruptio 1.31 0.75-2.30

Stratified by SGA
SGA AGA

OR 95% Cl OR 95% Cl
Active labor vs. no 1.93 0.99-3.73 1.25 0.81-1.95
active labor
Pre-eclampsia 0.47 0.18-1.21 0.40 0.15-1.10
Amnionitis 3.93 097-1597 2.15 1.03-4.48
Placenta abruptio 1.94 0.77-4.90 1.06 0.52-2.16

 

 

Multivariate models and stratified models are adjusted for birth weight,
gestational age, multiple gestation and other variables in the same model.

35

Table 3.8: Logistic rec

Model 1:

ression for PEL/VE:

 

C/S with latent labor
C/S with active labor
Vaginal delivery
Pre-eclampsia
Amnionitis

Placenta abruptio

C/S with latent labor
C/S with active labor
Vaginal delivery
Pre-eclampsia
Amnionitis

Placenta abruptio

Univariate model

Multivariate models

OR 95% Cl OR 95% CI

1 .48 0.65-3.35 1.42 0.57-3.57

2.65 1.24-5.65 2.61 1.09-6.25

1 .97 0.97-3.98 1.71 0.75-3.92
0.38 0.13-1.13
1.29 0.54-3.11
3.04 1.59-5.80

Stratified by SGA
SGA AGA

OR 95% Cl OR 95% Cl

2.55 0.47-13.96 1.15 0.37-3.54

5.36 103-2785 1.86 0.65539

3.71 073-1888 1.19 0.44-3.19

0.64 0.17-2.38 0.28 0.04-2.18

6.50 1 .45-29.18 0.39 0.20-2.41

2.10 0.65-6.80 3.61 1.62-8.05

 

Model 2.

 

Active labor vs. no
active labor
Pre-eclampsia
Amnionitis
Placenta abruptio

Active labor vs. no
active labor
Pre-eclampsia
Amnionitis

 

Placenta abruptio

Univariate model

Multivariate models

OR 95% CI OR 95% CI

1 .74 1 09-277 1.63 0.96-2.77
0.38 0.13-1.10
1.34 0.56-3.22
3.21 1.69-6.09

Stratified by SGA
SGA AGA

OR 95% CI OR 95% CI

2.45 0.97-6.16 1.28 0.68-2.44

0.61 0.17-2.28 0.28 0.04-2.14

6.90 1 57-3023 0.70 0.20-2.44

2.51 0.80-7.92 3.64 1.64-8.09

 

Multivariate models and stratified models are adjusted for birth weight,
gestational age, multiple gestation and other variables in the same model.

36

Table 3.9: Logistic regression for neonatal death:

 

Model 1:
Univariate model Multivariate models

OR 95% CI OR 95% Cl
C/S with latent labor 0.91 0.40-2.08 0.71 0.29-1.74
C/S with active labor 2.13 1.04-4.36 1.55 0.69-3.47
Vaginal delivery 2.85 1.51 -5.36 1.81 0.87-3.79
Pre-eclampsia 0.59 0.23-1 .51
Amnionitis 0.76 0.30-1.92
Placenta abruptio 1.42 0.71-2.86

Stratified by SGA
SGA AGA
OR 95% Cl OR 95% Cl

C/S with latent labor 2.61 057-1206 0.36 0.11-1.22
C/S with active labor 4.38 098-1954 0.88 0.33-2.38

 

 

 

 

Vaginal delivery 4.42 1.11-17.63 1.18 0.48-2.89
Pre-eclampsia 0.23 0.05-1 .07 1.64 0.47-5.68
Amnionitis 0.85 0.16-4.51 0.60 0.18-1.98
Placenta abruptio 1.62 0.54-4.83 1.10 0.41 -2.91
Model 2.
Univariate model Multivariate models
OR 95% Cl OR 95% Cl

Active labor vs. no 2.76 1.74-4.38 2.07 1 .21-3.53
active labor

Pre-eclampsia 0.61 0.24-1.55
Amnionitis 0.75 0.30-1 .90
Placenta abruptio 1.36 0.68-2.72

Stratified by SGA
SGA AGA
OR 95% CI OR 95% CI

Active labor vs. no 2.52 1.05-6.04 1.79 0.91-3.50
active labor

Pro-eclampsia 0.21 0.05-0.98 2.04 0.61-6.80
Amnionitis 0.86 0.17-4.40 0.62 0.19-2.04
Placenta abruptio 1.88 0.66-5.38 1.09 0.42-2.87

 

Multivariate models and stratified models are adjusted for birth weight,
gestational age, multiple gestation and other variables in the same model.

37

Table 3.10: Logistic re

Model 1:

ression for disabling CP:

 

C/S with latent labor
C/S with active labor
Vaginal delivery
Pre-eclampsia
Amnionitis

Placenta abruptio

C/S with latent labor
C/S with active labor
Vaginal delivery

Univariate model
OR 95% Cl
1.31 0.51-3.37
2.34 0.97-5.62
1 .46 0.64-3.33

Multivariate models
OR 95% CI
0.92 0.30-2.84
1.58 0.53-4.69
1.13 0.41-3.11

0.10 0.01-0.77
4.03 1 .49-10.87
5.27 2.32-12.01
Stratified by SGA
SGA AGA

OR 95% Cl OR 95% CI

3.76 0.41-34.57 0.97 0.32-2.92

4.07 0.36-46.57 1 .65 0.64-4.28

3.65 0.40-33.55

0.96 0.39-2.38

 

Model 2:

 

Active labor vs. no
active labor
Pre-eclampsia
Amnionitis
Placenta abruptio

Active labor vs. no
active labor

 

Univariate model
OR 95% Cl
1.49 0.84-2.62

Multivariate models
OR 95% Cl
1.39 0.70-2.69

0.10 0.01-0.81

4.11 153-1104
5.22 2.33-11.72
Stratified by SGA
SGA AGA
OR 95% CI OR 95% Cl

1 .56 0.46-5.26

1.19 0.62-2.28

 

#Multivariate models are adjusted for birth weight, gestational age, multiple
gestation and other variables in the same model. Stratified analysis is only done

in univariate models due to small sample size.

38

CHAPTER 4

DISCUSSION

This study evaluated the role of obstetric factors, focusing especially on
the presence of labor, on the development of neonatal death, ultrasound imaged
neonatal brain damages and DCP in LBW babies. Of the many obstetric
variables evaluated, several factors were found to influence the outcomes. As
expected, infants with adverse outcomes weighed less, and were less advanced
in gestation. Active labor, especially cesarean section with active labor, appeared
to have a significant influence on the risk of GM/IVH. Although the associations
were not very strong (OR=1.5-2.0), they reached statistical significance even
after controlling for potential confounders. The relationships of labor and PELNE
and neonatal death were clearly modified by SGA status. There was no modified
effect of SGA in the relationship of exposure to GM/IVH. In the overall
multivariate model, controlling for SGA did not change the effect of exposure on
GM/IVH. This phenomenon may be explained by small sample size in SGA
subgroup. There were 343 SGA and 715 AGA babies in the analysis. By roughly
testing the data using statistic simulation, around 700 subjects would have been
needed to obtain significance of effect of labor on GM/IVH in multivariate
analyses. Therefore we did not have enough power to get statistical significance
in subgroup (SGA) analysis. The effects of active labor on PEUVE and neonatal

death only existed in SGA infants. In AGA babies, these relationships

39

disappeared. Spontaneous preterm labor didn’t show increased risk of adverse
outcomes compared to iatrogenic preterm labor.

Regarding other obstetric factors, amnionitis showed a strong risk for
GM/IVH and DCP. The strong effect of amnionitis on PELNE only appeared in
SGA babies. Placenta abruptio was significantly associated with PEL/VE and
DCP. The effect on PEUVE only existed in AGA infants. Pre-eclampsia appeared
protective for DCP.

It is well known that the heightened risk of neurodevelopmental
abnormalities in low birth weight/preterm babies cannot be explained by single
factor, and may be caused by a combination of multiple risks. Although active
labor may only play a small role in the causal chain leading to adverse outcomes
in subgroup population, it is one of a limited number of potential risk factors that
may be controlled beforehand by obstetricians.

Among etiologic studies of brain damage in LBW babies, the mechanical
hypothesis of labor has a straightforward appeal. Uterine contractions reduce
uterine blood flow and oxygen delivery to the fetal-placental unit. It has also been
suggested that compression of the fetal head during labor and delivery might
alter intracranial capillary and venous pressures. The autoregulation of cerebral
blood flow in preterm babies does not occur consistently and is easily impaired
during episodes of hypoxia [80]. Perhaps active phase labor accelerates the
development of brain lesions in susceptible infants. Births delivered by cesarean
section usually are accompanied by maternal/fetal complications. It is possible
that active phase labor aggravates already compromised fetal oxygenation,

which may have resulted from conditions leading to cesarean sections.

40

SGA babies tend to be more mature than most babies at the same birth
weight. They are more likely to be free of brain damage if not complicated by
other risks. While intrauterine growth retardation (IUGR) causes a majority of
SGA, some infants are affected because of the limited growth potential inherited
from their parents. IUGR is commonly associated with maternal hypertension,
anemia, heart disease, and renal disease. Any factor that interferes with the
physiologic exchange in intervillous perfusion may cause IUGR. Growth-retarded
fetuses seem to have a diminished tolerance for hypoxia. The fetal oxygen
reserve is significantly reduced in these babies. They are more vulnerable to
strong labor. Maternal circulation is impaired in hypertensive mothers and IUGR
is frequently seen in pre-eclampsia cases. Pre-eclampsia itself showed a
protective effect for infant outcomes. However, this protective effect was diluted
by the damaging effects of active labor. In tables 3.2-3.5, when active labor was
combined with other risks (e.g. amnionitis or placenta abruptio), the rate of
adverse outcomes increased remarkably. If the mechanical energy of active labor
itself is not strong enough to cause brain damage in normal preterm babies, it
might be sufficient to initiate brain damage and death when combined with the a
reduced fetal oxygen reserve.

There is no effect of spontaneous preterm labor on the outcomes. The
duration and severity of the asphyxial period is difficult to assess, even with the
advent of electronic fetal heart monitoring and the measurement of scalp and
cord pH and base deficits [81]. In severely compromised fetuses, active labor
may initiate brain damage even at levels of oxygenation that fail to evoke any

detectable general fetal circulatory response.

4|

The cardinal feature of brain damage in preterm infants is injury to
hemispheric white matter. The white matter, containing the preponderance of
motor projection fiber bundles, controls motor function in the limbs. It is known
that ultrasonongraphic evidence of damage to white matter in the neonatal period
is the most important predictor of CP in preterm infants. These lesions are

responsible for the concept of “grade IVhemorrhage” and viewed as an extension

of IVH in some studies. In NBH, these lesions were represented as PELNE
overall. PELNE (especially echodensity) usually can be seen in ultrasound scan
at 7 days, although they are fully developed at several weeks of life. In NBH
cohort, most of the PELNE were detected by early scans. Later scans (11 days
to discharge) only accounted for a small proportion (14%) of newly detected
PELNE.

Cesarean section with active labor accounted for a moderately strong risk
of PEL/VE in SGA babies. But active labor was not linked to an increased risk for
DCP. This phenomenon can be explained by three ways. Child death may be the
cause of this discrepancy. The effect of cesarean section with active labor on
PEUVE is much stronger in children who died before 2 years follow-up than in
survivors. The brain lesions associated with active labor may represent the
severest cases. Active labor may disproportionately affect the sickest fetuses by
causing death rather than handicap. Another explanation is active labor could
cause structural anomalies of PEUVE. It was known that almost all
ultrasonographic echolucencies resulted in GP, but echodensities were not
always associated with CP. Hemorrhagic white matter damage is more lethal

and seen more in early ultrasound. The small subgroup (SGA) sample size may

42

not have enough power to detect the significance of the association, especially
after two years follow-up.

Vaginal delivery had the highest risk compared to cesarean section with
no labor for neonatal death in univariate but not in multivariate analyses.
Generally in clinical practice, the lowest birth weight babies are more commonly
delivered vaginally. Obstetricians don’t want to do a cesarean section on these
cases. On the other hand, these babies come out quickly. The effect of vaginal
delivery on neonatal death may therefore be confounded by birth weight.

Some neonatal factors (e.g. Apgar scores, CLD) are related to neonatal
outcomes, especially later impairments. They may play roles as mediators of the
relationship between active labor and outcomes.

The NBH dataset contains detailed information on the perinatal period of a
large cohort. It has enough sample size to assess intrapartum risks and adverse
outcomes. A well-designed protocol for ultrasound screening in neonates
covered nearly all the high yield periods for neonatal brain lesions. High follow up
rates for developmental outcomes provide good data for learning the etiology and
risks in low birth weight babies. Some shortcomings exist in this dataset.
Obtaining obstetric information from hospital charts has its limitations. Time and
duration of labor phases usually can be abstracted correctly from the chart.
Although cervical dilatation measurement may vary between medical personal,
only a small portion of our exposures was defined purely by cervical dilatation
and thus results should not be greatly affected. Matemal/fetal complications
abstracted from chart diagnosis may be affected by the judgment of physicians.

The major potential confounder, pre-eclampsia, was defined by blood pressure

43

measurement and medical history rather than purely from the medical diagnosis.
Moreover, we didn’t have enough cases using multiple regression models to test
the effect of active labor in amnionitis or placenta abruptio patients. But these
limitations should not critically affect the study results.

In closing, the findings reported here showed active labor itself did not
have increased risk for neonatal death and neurodevelopmental outcomes on
normal infants. However, when active labor was imposed on already
compromised fetus (e.g. SGA), the energy of active labor may impair intrauterine
environment and cause adverse outcomes. Other factors (e.g. amnionitis,
placenta abruptio) that compromise the fetal oxygen reserve may have the same
interaction with labor as SGA. The clinical importance of active labor was mainly
found in mortality and brain damage, but not childhood handicap. In order to
obtain more definitive conclusions, clinical trials should be considered though
difficult to perform. The clinical decisions about labor and delivery in low birth
weight babies should be based on balancing potential advantages and

adversities on both mothers and babies.

44

10.

REFERENCES

MacDorman, M.F. and J.O. Atkinson, Infant mortality statistics from the
linked birth/infant death data set--1995 period data. Mon Vital Stat Rep,
1998. 46(6 Suppl 2): p. 1-22.

Ellenberg, J.H. and KB. Nelson, Birth weight and gestational age in
children with cerebral palsy or seizure disorders. American Journal Of
Diseases Of Children, 1979. 133(10): p. 1044-8.

Mervis, C.A., et al., Low birthweight and the risk for mental retardation
later in childhood. Paediatr Perinat Epidemiol, 1995. 9(4): p. 455-68.

Crofts, B.J., R. King, and A. Johnson, The contribution of low birth weight
to severe vision loss in a geographically defined population [see
comments]. Br J Ophthalmol, 1998. 82(1): p. 9-13.

Veen, S., et al., Hearing loss in very preterm and very low birthweight
infants at the age of 5 years in a nationwide cohort. International Journal
Of Pediatric Otorhinolaryngology, 1993. 26(1): p. 11-28.

Lanska, M.J. and DJ Lanska, Neonatal seizures in the United States:
results of the National Hospital Discharge Survey, 1980- 1991.
Neuroepidemiology, 1996. 15(3): p. 117-25.

Hack, M., et al., The effect of very low birth weight and social risk on
neurocognitive abilities at school age. Journal Of Developmental And
Behavioral Pediatrics, 1992. 13(6): p. 412-20.

deRegnier, RA, et al., Association between the severity of chronic lung
disease and first-year outcomes of very low birth weight infants. Journal
Of Perinatology, 1997. 17(5): p. 375-82.

Escobar, G.J., B. Littenberg, and DB. Petitti, Outcome among surviving
very low birthweight infants: a meta-analysis. Archives Of Disease In
Childhood, 1991. 66(2): p. 204-11.

Veen, S., et al., Impairments, disabilities, and handicaps of very preterm
and very-Iow-birth weight infants at five years of age. The Collaborative
Project on Preterm and Small for Gestational Age Infants (POPS) in The
Netherlands [see comments]. Lancet, 1991. 338(8758): p. 33-6.

45

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Omstein, M., et al., Neonatal follow-up of very low birthweight/extremely
low birthweight infants to school age: a critical overview. Acta Paediatrica
Scandinavica, 1991. 80(8-9): p. 741-8.

Preterm singleton births--United States, 1989-1996. mer. Morbidity And
Mortality Weekly Report, 1999. 48(9): p. 185-9.

Paneth, N., Recent trends in preterm delivery rates in the United States.
Prevention of Preterm Birth. Inserrn, 1986(138): p. 15-30.

Ventura, S.J., et al., Births: final data for 1997. Natl Vital Stat Rep, 1999.
47(18): p. 1-96.

Paneth, NS, The problem of low birth weight. future of children, 1995.
5(1): p. 19-34.

Goldenberg, R.L. and DJ Rouse, Prevention of premature birth [see
comments]. New England Journal of Medicine, 1998. 339(5): p. 313-20.

Alexander, G.R., et al., Preterm birth prevention: an evaluation of
programs in the United States. Birth, 1991. 18(3): p. 160-9.

Paneth, N., Neonatal care and patterns of handicap in the community.
Clinic Dev Med, 1992. 123(124): p. 232-241.

Guyer, B., et al., Annual summary of vital statistics-1997. Pediatrics, 1998.
102(6): p. 1333-49.

Lorenz, J.M., et al., A quantitative review of mortality and developmental
disability in extremely premature newborns. archives of pediatrics and
adolescent medicine, 1998. 152(5): p. 425-35.

Bhushan, V., N. Paneth, and J.L. Kiely, Impact of improved survival of
very low birth weight infants on recent secular trends in the prevalence of
cerebral palsy [see comments]. pediatrics, 1993. 91(6): p. 1094-100.

Pape, K.E., et al., Ultrasound detection of brain damage in preterm
infants. Lancet, 1979. 1(8129): p. 1261-4.

Levene, M.l., J.S. Wigglesworth, and V. Dubowitz, Hemorrhagic
periventricular Ieukomalacia in the neonate: a real-time ultrasound study.
Pediatrics, 1983. 71 (5): p. 794-7.

Pinto-Martin, J., et al., The central New Jersey neonatal brain

haemorrhage study: design of the study and reliability of ultrasound
diagnosis. paediatric and perinatal epidemiology, 1992. 6(2): p. 273-84.

46

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

Graziani, L.J., et al., Cranial ultrasound and clinical studies in preterm
infants. Journal of Pediatrics, 1985. 106(2): p. 269-76.

Scott, D.T., et al., Evidence for late developmental deficit in very low birth
weight infants surviving intraventricular hemorrhage. Childs Brain, 1984.
11(4): p. 261-9.

Palmer, P., et al., Developmental and neurological progress of preterm
infants with intra ventricular haemorrhage and ventricular dilatation.
Archives of Disease in Childhood, 1982. 57(10): p. 748-53.

Sostek, A.M., et al., Developmental outcome of preterm infants with
intraventricular hemorrhage at one and two years of age. Child
Development, 1987. 58(3): p. 779-86.

Pinto-Martin, J.A., et al., Cranial ultrasound prediction of disabling and
nondisabling cerebral palsy at age two in a low birth weight population.
pediatrics, 1995. 95(2): p. 249-54.

Paneth, N., et al., White matter necrosis in very low birth weight infants:
neuropathologic and ultrasonographic findings in infants surviving six days
or longer. journal of pediatrics, 1990. 116(6): p. 975-84.

Guzzetta, F., et al., Periventricular intraparenchymal echodensities in the
premature newbom: critical determinant of neurologic outcome. Pediatrics,
1986. 78(6): p. 995-1006.

Fawer, C.L., A. Calame, and MT. Furrer, Neurodevelopmental outcome at
12 months of age related to cerebral ultrasound appearances of high risk
preterm infants. Early Human Development, 1985. 11(2): p. 123-32.

Dammann, O. and A. Leviton, Brain damage in preterm newboms:
biological response modification as a strategy to reduce disabilities.
Journal of Pediatrics, 2000. 136(4): p. 433-8.

Kuban, K., et al., White matter disorders of prematurity: association with
intra ventricular hemorrhage and ventriculomegaly. The Developmental
Epidemiology Network [see comments]. Journal of Pediatrics, 1999.
134(5): p. 539-46.

Graziani, L.J., et al., Neonatal neurosonographic correlates of cerebral
palsy in preterm infants. Pediatrics, 1986. 78(1): p. 88-95.

Graham, M., et al., Prediction of cerebral palsy in very low birthweight
infants: prospective ultrasound study. Lancet, 1987. 2(8559): p. 593-6.

47

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

Beverley, D.W., G.W. Chance, and CF. Coates, lntraventricular
haemorrhage--timing of occurrence and relationship to perinatal events.
British Journal of Obstetrics and Gynaecology, 1984. 91 (10): p. 1007-13.

de Crespigny, L.C., et al., Timing of neonatal cerebroventricular
haemorrhage with ultrasound. Archives of Disease in Childhood, 1982.
57(3): p. 231-3.

Murton, L.J., et al., Perinatal factors, periventricular haemorrhage and
mortality in very low birthweight infants. Australian Paediatric Journal,
1985. 21(1): p. 39-43.

Paneth, N., et al., Incidence and timing of germinal matrix/intraventricular
hemorrhage in low birth weight infants. american journal of epidemiology,
1993. 137(11): p. 1167-76.

Little, W.l., On the influence of abnormal parturition, difficult labours,
premature birth, and asphyxia neonatorum, on the mental and physical
conditions of the child, especially in relation to deformities. Transactions of
the Obstetrical Society of London. Vol. 3. 1862. 293-344.

Freud, S., Infantile Cerebral Paralysis. 1897: University of Miami Press.

Caldeyro-Barcia, Physiology of the uterine contraction. Clin Obstet Gynco,
1960.

Avery, G.B., Noenatology: pathophysiology and management of the
newborn. 1981: Lippincott.

Jacobson, L. and G. Rooth, Interpretative aspects on the acid-base
composition and its variation in fetal scalp blood and maternal blood
during labour. Journal of Obstetrics and Gynaecology of the British
Commonwealth, 1971. 78(11): p. 971 -80.

Lipscomb, A.P., et al., Pneumothorax and cerebral haemorrhage in
preterm infants. Lancet, 1981. 1(8217): p. 414-6.

Cowan, F. and M. Thoresen, Changes in superior sagittal sinus blood
velocities due to postural alterations and pressure on the head of the
newborn infant. Pediatrics, 1985. 75(6): p. 1038-47.

Fujimura, M., et al., Clinical events relating to intraventricular
haemorrhage in the newborn. Archives of Disease in Childhood, 1979.
54(6): p. 409-14.

Reynolds, ML, et al., lntracranial haemorrhage in the preterm sheep
fetus. Early Human Development, 1979. 3(2): p. 163-86.

48

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

Garfield, R.E., Control of myometrial function in preterm versus term labor.
Clinical Obstetrics and Gynecology, 1984. 27(3): p. 572-91.

Harris, B.A.J., H. Gore, and C.E.J. Flowers, Peripheral placental
separation: a possible relationship to premature labor. Obstetrics and
Gynecology, 1985. 66(6): p. 774-8.

Hameed, C., et al., Silent chorioamnionitis as a cause of preterm labor
refractory to tocolytic therapy. American Journal of Obstetrics and
Gynecology, 1984. 149(7): p. 726-30.

Leviton, A., et al., Labor and delivery characteristics and the risk of
germinal matrix hemorrhage in low birth weight infants. Journal Of Child
Neurology, 1991. 6(1): p. 35-40.

Hansen, A. and A. Leviton, Labor and delivery characteristics and risks of
cranial ultrasonographic abnormalities among very-low-birth-weight
infants. The Developmental Epidemiology Network Investigators.
American Journal Of Obstetrics And Gynecology, 1999. 181(4): p. 997-

1 006.

Anderson, G.D., et al., The relationship between labor and route of
delivery in the preterm infant. American Journal Of Obstetrics And
Gynecology, 1988. 158(6 Pt 1): p. 1382-90.

Anderson, G.D., et al., The effect of cesarean section on intraventricular
hemorrhage in the preterm infant. American Journal Of Obstetrics And
Gynecology, 1992. 166(4): p. 1091-9; discussion 1099-101.

Shaver, D.C., et al., Early and late intraventricular hemorrhage: the role of
obstetric factors. Obstetrics And Gynecology, 1992. 80(5): p. 831 -7.

Tejani, N., et al., Method and route of delivery in the low birth weight
vertex presentation correlated with early periventricular/intra ventricular
hemorrhage. Obstetrics And Gynecology. 1987. 69(1): p. 1-4.

Malloy, M.H., L. Onstad, and E. Wright, The effect of cesarean delivery on
birth outcome in very low birth weight infants. National Institute of Child
Health and Human Development Neonatal Research Network. Obstetrics
And Gynecology, 1991. 77(4): p. 498-503.

Wallace, R.L., B.S. Schifrin, and RH. Paul, The delivery route for very-

Io w-birth-weight infants. A preliminary report of a randomized, prospective
study. Journal Of Reproductive Medicine, 1984. 29(10): p. 736-40.

49

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

Westgren, M., et al., Mode of delivery in the low birth weight fetus.
Delivery by cesarean section independent of fetal lie versus vaginal
delivery in vertex presentation. A study with long-term follow-up. Acta
Obstetricia Et Gynecologica Scandinavica, 1985. 64(1): p. 51 -7.

Ment, L.R., et al., Antenatal steroids, delivery mode, and intraventricular
hemorrhage in preterm infants. American Journal Of Obstetrics And
Gynecology. 1995. 172(3): p. 795-800.

Baud, O., et al., Are neonatal brain lesions due to intrauterine infection
related to mode of delivery? British Journal Of Obstetrics And
Gynaecology, 1998. 105(1): p. 121-4.

Jonas, HA and J.M. Lumley, The effect of mode of delivery on neonatal
mortality in very low birthweight infants born in Victoria, Australia:
Caesarean section is associated with increased survival in breech-
presenting, but not vertex-presenting, infants. Paediatric and Perinatal
Epidemiology, 1997. 11(2): p. 181-99.

Gbrbe, E., et al., Very-low-birthweight breech infants: short-terrn outcome
by method of delivery. Journal Of Matemal-Fetal Medicine, 1997. 6(3): p.
155-8.

Jain, L., C. Ferre, and D. Vidyasagar, Cesarean delivery of the breech
very-low-birth-weight infant: does it make a difference? Journal Of
Matemal-Fetal Medicine, 1998. 7(1): p. 28-31.

Lee, K.S., et al., Relationship of cesarean delivery to lower birth weight-
specific neonatal mortality in singleton breech infants in the United States
[see comments]. Obstetrics and Gynecology, 1998. 92(5): p. 769-74.

O'Shea, M., et al., Prenatal events and the risk of
subependymaI/intra ventricular haemorrhage in very low birthweight
neonates. Paediatric And Perinatal Epidemiology, 1992. 6(3): p. 352-62.

Leviton, A., et al., Matemal receipt of magnesium sulfate does not seem to
reduce the risk of neonatal white matter damage. pediatrics, 1997. 99(4):
p. E2.

Spinillo, A., et al., Preeclampsia, preterm delivery and infant cerebral
palsy. European Journal Of Obstetrics, Gynecology, And Reproductive
Biology, 1998. 77(2): p. 151-5.

Grether, J.K., et al., Prenatal and perinatal factors and cerebral palsy in
very low birth weight infants [see comments]. Journal Of Pediatrics, 1996.
128(3): p. 407-14.

50

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

Murphy, D.J., et al., Case-control study of antenatal and intrapartum risk
factors for cerebral palsy in very preterm singleton babies. Lancet, 1995.
346(8988): p. 1449-54.

Torts, C.P., et al., Prenatal and perinatal factors in the etiology of cerebral
palsy [see comments]. Journal of Pediatrics, 1990. 116(4): p. 615-9.

Papile, L.A., et al., Incidence and evolution of subependymal and
intra ventricular hemorrhage: a study of infants with birth weights less than
1,500 gm. Journal of Pediatrics, 1978. 92(4): p. 529-34.

Holzman, GB, The effects of host factors and prenatal alcohol exposure
on the risk of perinatal brain injury in low birth weight infants. 1996.

Baylay, N., Bayley Scales of Infant Development. 1969, New York:
Psychological Corporation.

Holzman, C., et al., Perinatal brain injury in premature infants born to
mothers using alcohol in pregnancy. Neonatal Brain Hemorrhage Study
Team. pediatrics, 1995. 95(1): p. 66-73.

Yudkin, P.L., et al., New birthweight and head circumference centiles for
gestational ages 24 to 42 weeks. Early Human Development, 1987. 15(1):
p. 45-52.

Paneth, N., et al., Magnesium sulfate in labor and risk of neonatal brain
lesions and cerebral palsy in low birth weight infants. The Neonatal Brain
Hemorrhage Study Analysis Group. pediatrics, 1997. 99(5): p. E1.

Lou, H.C., N.A. Lassen, and B. Friis-Hansen, Impaired autoregulation of
cerebral blood flow in the distressed newborn infant. Journal of Pediatrics,
1979. 94(1): p. 118-21.

Paneth, N. and R.l. Stark, Cerebral palsy and mental retardation in

relation to indicators of perinatal asphyxia. An epidemiologic overview.
american journal of obstetrics and gynecology, 1983. 147(8): p. 960-6.

51