Original Research ajog.

org

OBSTETRICS
Mortality and pulmonary outcomes of extremely preterm
infants exposed to antenatal corticosteroids
Colm P. Travers, MD; Waldemar A. Carlo, MD; Scott A. McDonald, BS; Abhik Das, PhD; Edward F. Bell, MD;
Namasivayam Ambalavanan, MD; Alan H. Jobe, MD, PhD; Ronald N. Goldberg, MD; Carl T. D’Angio, MD;
Barbara J. Stoll, MD; Seetha Shankaran, MD; Abbot R. Laptook, MD; Barbara Schmidt, MD, MSc;
Michele C. Walsh, MD; Pablo J. Sánchez, MD; M. Bethany Ball, BSc, CCRC; Ellen C. Hale, RN, BSc, CCRC;
Nancy S. Newman, RN; Rosemary D. Higgins, MD; for the Eunice Kennedy Shriver National Institute of Child Health
and Human Development Neonatal Research Network

BACKGROUND: Antenatal corticosteroids are given primarily to interval, 0.70e0.85; P < .0001). In an analysis by each week of
induce fetal lung maturation but results from meta-analyses of randomized gestation, infants exposed to a complete course of antenatal corti-
controlled trials have not shown mortality or pulmonary benefits for costeroids had lower mortality before discharge compared to infants
extremely preterm infants although these are the infants most at risk of without exposure at each week from 23-27 weeks’ gestation and
mortality and pulmonary disease. infants exposed to a partial course of antenatal corticosteroids had
OBJECTIVE: We sought to determine if exposure to antenatal corti- lower mortality at 23, 24, and 26 weeks’ gestation. Rates of bron-
costeroids is associated with a lower rate of death and pulmonary mor- chopulmonary dysplasia in survivors did not differ by antenatal
bidities by 36 weeks’ postmenstrual age. corticosteroid exposure. The rate of death due to respiratory distress
STUDY DESIGN: Prospectively collected data on 11,022 infants 22 syndrome, the rate of surfactant use, and the rate of mechanical
0/7 to 28 6/7 weeks’ gestational age with a birthweight of
401 g born ventilation were lower in infants exposed to any antenatal cortico-
from Jan. 1, 2006, through Dec. 31, 2014, were analyzed. The rate of steroids compared to infants without exposure.
death and the rate of physiologic bronchopulmonary dysplasia by 36 CONCLUSION: Among infants 22-28 weeks’ gestational age, any or
weeks’ postmenstrual age were analyzed by level of exposure to antenatal partial antenatal exposure to corticosteroids compared to no exposure is
corticosteroids using models adjusted for maternal variables, infant associated with a lower rate of death while the rate of bronchopulmonary
variables, center, and epoch. dysplasia in survivors did not differ.
RESULTS: Infants exposed to any antenatal corticosteroids had a
lower rate of death (2193/9670 [22.7%]) compared to infants without Key words: antenatal corticosteroids, bronchopulmonary dysplasia,
exposure (540/1302 [41.5%]) (adjusted relative risk, 0.71; 95% infant, intracranial hemorrhage, mechanical ventilation, morbidity, mor-
confidence interval, 0.65e0.76; P < .0001). Infants exposed to a tality, necrotizing enterocolitis, neonatal, newborn, patent ductus arte-
partial course of antenatal corticosteroids also had a lower rate of riosus, periventricular leukomalacia, pneumothorax, preterm, pulmonary,
death (654/2520 [26.0%]) compared to infants without exposure pulmonary hemorrhage, respiratory distress syndrome, respiratory sup-
(540/1302 [41.5%]); (adjusted relative risk, 0.77; 95% confidence port, sepsis, surfactant

Introduction meta-analyses of antenatal corticoste- together and separately. Antenatal

The effects of antenatal corticosteroids roids show no reduction in respiratory corticosteroid exposure may affect both
on mortality and pulmonary outcomes distress syndrome or neonatal death for outcomes for example if more infants
at the lowest gestations show mixed infants delivered <30 weeks’ gesta- survive following exposure and then
results, in part due to the small sample tion.1,2 In addition, there are limited develop bronchopulmonary dysplasia
size of randomized controlled trials.1,2 data from observational studies subsequently.
Although antenatal corticosteroids are comparing the pulmonary outcomes of A complete course of antenatal corti-
given primarily to induce pulmonary extremely preterm infants exposed to costeroids is defined as 2 intramuscular
maturity, induce surfactant release, and antenatal corticosteroids to those doses of betamethasone given 12-24
decrease respiratory distress syndrome, without exposure because these studies hours apart or 4 intramuscular doses of
randomized controlled trials and have not been focused on pulmonary dexamethasone given 12 hours apart.10
outcomes.3-8 Extremely preterm infants Many preterm infants are born prior to
who die <36 weeks’ postmenstrual age the administration of a complete course
Cite this article as: Travers CP, Carlo WA, McDonald SA, cannot be assessed for the development of antenatal corticosteroids.11,12 There
et al. Mortality and pulmonary outcomes of extremely of bronchopulmonary dysplasia, a type are insufficient data on mortality and
preterm infants exposed to antenatal corticosteroids. Am
of chronic lung disease that is pulmonary outcomes of extremely
J Obstet Gynecol 2018;218:130.e1-13.
diagnosed at 36 weeks’ postmenstrual preterm infants born after exposure to
0002-9378/$36.00 age.9 It is important to evaluate the either a complete or a partial course of
ª 2017 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.ajog.2017.11.554 competing outcomes of broncho- antenatal corticosteroids. We hypothe-
pulmonary dysplasia and death both sized that the rates of death would be

130.e1 American Journal of Obstetrics & Gynecology JANUARY 2018

ajog.org OBSTETRICS Original Research

based on respiratory support at 36 weeks’

FIGURE 1
postmenstrual age using the physiologic
Frequency of antenatal corticosteroids exposure by gestational age and year
definition, which uses an oxygen reduc-
tion challenge test among eligible in-
fants.15 The physiologic definition has
been shown to be more reliable and pre-
cise than the clinical definition of bron-
chopulmonary dysplasia15 (defined as
supplemental oxygen at 36 weeks’ post-
menstrual age) and has been used by the
Neonatal Research Network since 2006.
All other outcomes were based on stan-
dardized definitions as per the generic
database of the Neonatal Research
Network.16 Cause of death was defined as
the underlying proximate disease that
initiated the series of events leading to
death based on both clinical evidence and
autopsy findings where available.13
Frequency of exposure to antenatal corticosteroids (ANS) by gestational age and year of birth.
Administration of ANS increased over study period but remained lower at lower gestational ages.
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018.
Statistical analysis
The primary outcome measure was death
before discharge. A formal sample size
and power estimate demonstrated that
lower in infants exposed to antenatal resuscitated as these infants were less the sample size resulting from inclusion
corticosteroids. In addition, we hypoth- likely to have lethal anomalies. Infants of all infants delivered from Jan. 1, 2006,
esized that the rates of physiologic who died in the first 12 hours after birth through Dec. 31, 2014, would provide
bronchopulmonary dysplasia or death without delivery room resuscitation >95% power to detect an absolute dif-
would be lower in infants exposed to were excluded from the primary analysis ference of 4% centered around an overall
antenatal corticosteroids. This study was to ensure that results were not affected event rate of 25%. All secondary outcome
also designed to determine if exposure to by planned restriction of care but were measures and analyses were prespecified.
a partial or a complete course of included in a secondary analysis. The All outcomes were analyzed by level of
antenatal corticosteroids is associated study protocol was approved by each exposure to antenatal corticosteroid:
with improved survival and pulmonary center’s institutional review board. complete exposure, partial exposure, any
outcomes in extremely preterm infants. ClincalTrials.gov identifiers are: (partial or complete) exposure, and no
NCT00063063 (generic database) and exposure. Differences in categorical var-
Materials and Methods NCT00009633 (follow-up study). iables were described using Fisher exact
This was a hypothesis-driven study using test. Kruskal-Wallis test was used for
data collected prospectively for the Definitions continuous skewed variables. Robust
Neonatal Research Network Generic Infants were considered exposed to Poisson regression analysis was per-
Database and follow-up studies. These antenatal corticosteroids if their mother formed for factors present at birth asso-
data included infants 22 0/7 to 28 6/7 had received
1 doses of either ciated with pulmonary outcomes
weeks’ gestation with a birthweight of betamethasone or dexamethasone.10 including birthweight, sex, multiple

401 g born from Jan 1, 2006, Mothers were considered to have births, small for gestational age (<10th
through Dec. 31, 2014, at any of the received a complete course if they had centile), maternal variables (age, marital
Eunice Kennedy Shriver National Insti- received at least 2 doses and 24 hours had status, race, diabetes, rupture of mem-
tute of Child Health and Human passed from the time the first dose of branes
24 hours, antepartum hemor-
Development (NICHD) Neonatal antenatal corticosteroids was given. Data rhage, and mode of delivery), center, and
Research Network centers. Maternal and on repeat courses of antenatal cortico- epoch (2006 through 2009, and 2010
neonatal sociodemographic and clinical steroids were not collected.14 Data were through 2014).17-19 There were 0.5%
data were collected from medical records collected using standardized definitions missing data for the primary outcome
by trained research personnel. Gesta- until death or discharge. Follow-up data and 9.7% missing data for the follow-up
tional age was determined by best were collected using standardized outcomes at 18-22 months corrected
obstetric estimate over best neonatal definitions on eligible surviving infants at gestational age. To ensure that results
estimate.13 Infants with congenital 18-22 months corrected gestational age. were not affected by missing data,
anomalies were included if they were Bronchopulmonary dysplasia was defined multiple imputation analyses were

JANUARY 2018 American Journal of Obstetrics & Gynecology 130.e2

Original Research OBSTETRICS ajog.org

the analysis by each week of gestation,

TABLE 1
infants exposed to a complete course of
Infant/maternal baseline characteristics by antenatal corticosteroid
antenatal corticosteroids had lower
treatment
mortality before discharge compared to
Any ANS No ANS infants without exposure at each week
Study population, n 9715 1307 from 23-27 weeks’ gestation (Figure 2
and Table 3). Infants exposed to a
Mother
partial course of antenatal corticoste-
Race/ethnicity roids had lower mortality at 23, 24, and
Black, including black Hispanic 3893 (40.9)a 663 (52.0)a 26 weeks’ gestation (Figure 2 and
White, including white Hispanic 5066 (53.3) a
546 (42.8)a Table 3).
Infants exposed to any antenatal
Other 550 (5.8)a 66 (5.2)a
corticosteroids had a lower rate of
All Hispanic 1366 (14.5)a 288 (22.7)a physiologic bronchopulmonary
Health insurance dysplasia or death by 36 weeks’ post-
Medicaid 5096 (52.8)a 807 (62.4)a
menstrual age (6016/9579 [62.8%])
compared to infants without exposure
Private insurance 4072 (42.2)a 355 (27.5)a (940/1300 [72.3%]) (ARR, 0.94; 95% CI,
Self-pay/uninsured 480 (5.0)a 131 (10.1)a 0.91e0.98; P ¼ .001) (Table 2). The rate
Infant of physiologic bronchopulmonary
dysplasia or death by 36 weeks’ post-
Birthweight, g 748  148a 719  155a
menstrual age did not differ among
Gestational age, wk 25.5  1.5a 24.8  1.7a infants exposed to a partial course of
Male sex 4731 (48.7) 668 (51.1) antenatal corticosteroids compared to
Multiple births 2565 (26.4) a
285 (21.8)a infants without exposure
(Supplementary Table 1). The rate of
Small for gestational age 1078 (11.1)a 99 (7.6)a
physiologic bronchopulmonary
a
Caesarean delivery 6697 (68.9) 743 (56.9)a dysplasia among survivors at 36 weeks’
Values are n (%) or  SD. postmenstrual age did not differ signifi-
ANS, antenatal corticosteroid. cantly in infants exposed to any antenatal
a
Significant with P value <.05 for comparisons of ANS vs no ANS data. corticosteroids compared to infants
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018. without exposure (Table 2). The rate of
death due to bronchopulmonary
dysplasia did not differ significantly in
additionally conducted for outcomes antenatal corticosteroids increased over infants exposed to any antenatal corti-
with >1% of data missing, or when there the study period (Figure 1). Mothers of costeroids (172/9661 [1.8%]) compared
was an imbalance of missing data of infants exposed to antenatal corticoste- to infants without exposure (16/1299
>0.3% between groups. An additional roids were more likely to be white, [1.2%]) (ARR, 1.65; 95% CI, 0.96e2.83;
analysis was also performed adjusting delivered by cesarean delivery, with pri- P ¼ .068].
for chorioamnionitis (diagnosed by vate health insurance. Infants exposed to The rate of respiratory distress
placental pathology). This separate anal- antenatal corticosteroids had higher syndrome did not differ in infants
ysis was not performed by each week of birthweight, had longer gestational age, exposed to any antenatal corticosteroids
gestation. Software (SAS, Version 9.3; and were more likely to be small for compared to infants without exposure
SAS Institute Inc, Cary, NC) was used for gestational age and the product of (Table 2). The rate of death due to
all statistical analyses. Odds ratios and multiple births (Table 1). respiratory distress syndrome was lower
95% confidence intervals (CI) were esti- Infants exposed to any antenatal in infants exposed to any antenatal
mated for binary outcomes with a corticosteroids had lower mortality corticosteroids compared to infants
2-sided P value of <.05 indicating before discharge (2193/9670 [22.7%]) without exposure (Table 2). The rate of
statistical significance. compared to infants without exposure death due to respiratory distress syn-
(540/1302 [41.5%]) (adjusted relative drome was also lower in infants exposed
Results risk [ARR], 0.71; 95% CI, 0.65e0.76; to a complete or partial course of ante-
A total of 11,022 infants met the inclu- P < .0001) (Table 2). Infants exposed to natal corticosteroids compared to in-
sion criteria of whom 9715 (88.1%) were a partial course of antenatal corticoste- fants without exposure (Supplementary
exposed to antenatal corticosteroids. roids also had lower mortality before Table 1). The rate of surfactant use and
The proportion of infants at each gesta- discharge compared to infants without the rate of mechanical ventilation were
tional age from 23-28 weeks exposed to exposure (Supplementary Table 1). In lower in infants exposed to any antenatal

130.e3 American Journal of Obstetrics & Gynecology JANUARY 2018

ajog.org OBSTETRICS Original Research

TABLE 2
Outcomes of infants by exposure to antenatal corticosteroids
Any ANS/total, n (%) No ANS/total, n (%) ARR (95% CI)a
Total study population N ¼ 9715 N ¼ 1307
Death
By 36 wk’ postmenstrual age 1952/9692 (20.1) 513/1305 (39.3) 0.67 (0.62e0.73)b
Before discharge 2193/9670 (22.7) 540/1302 (41.5) 0.71 (0.65e0.76)b
Due to bronchopulmonary dysplasia 172/9661 (1.8) 16/1299 (1.2) 1.65 (0.96e2.83)c
Due to respiratory distress syndrome 698/9661 (7.2) 171/1299 (13.2) 0.72 (0.60e0.86)b,c
Bronchopulmonary dysplasia [physiologic definition] 6016/9579 (62.8) 940/1300 (72.3) 0.94 (0.91e0.98)b
or death by 36 wk’ postmenstrual age, 2006 through 2014
Population of survivors N ¼ 7477 N ¼ 762
Bronchopulmonary dysplasia, physiologic definition 3810/7359 (51.8) 396/755 (52.5) 0.96 (0.89e1.03)
Bronchopulmonary dysplasia, by use of supplemental 3999/7431 (53.8) 415/759 (54.7) 0.96 (0.90e1.03)
oxygen at 36 wk’ postmenstrual age, clinical definition
Respiratory distress syndrome 7367/7477 (98.5) 760/762 (99.7) 0.99 (0.99e1.00)
Surfactant use 6347/7477 (84.9) 702/762 (92.1) 0.92 (0.89e0.94)b
Mechanical ventilation 6742/7472 (90.2) 723/762 (94.9) 0.96 (0.94e0.98)b
Pneumothorax 354/7477 (4.7) 45/762 (5.9) 0.78 (0.56e1.08)c
Pulmonary hemorrhage 305/7477 (4.1) 47/762 (6.2) 0.75 (0.55e1.03)c
Treatment with postnatal steroids for 1218/7268 (16.8) 112/733 (15.3) 0.98 (0.82e1.18)
bronchopulmonary dysplasia
Early-onset sepsis 134/7477 (1.8) 17/762 (2.2) 0.67 (0.40e1.13)c
Necrotizing enterocolitis stage
2 668/7476 (8.9) 71/762 (9.3) 0.98 (0.77e1.26)c
Intracranial hemorrhage/periventricular leukomalacia 1028/7445 (13.8) 167/761 (21.9) 0.66 (0.57e0.77)b,c
Patent ductus arteriosus treated with 2454/7471 (32.8) 293/761 (38.5) 0.95 (0.86e1.05)
indomethacin/ibuprofen
Retinopathy of prematurity stage
3 or treated 1325/7342 (18.0) 178/753 (23.6) 0.76 (0.66e0.87)b,c
with ablation/anti-VEGF drug
Respiratory support at discharge, oxygen 2386/7216 (33.1) 195/740 (26.4) 1.17 (1.03e1.33)b,c
Prolonged hospital stay
120 d, all causes 2086/7315 (28.5) 235/752 (31.3) 0.94 (0.84e1.05)
Population of survivors eligible for follow-up N ¼ 4149 N ¼ 471
Respiratory support at 18e22 mo corrected 67/3749 (1.8) 7/422 (1.7) 0.91 (0.42e1.99)c
age, ventilation/CPAP
Oxygen at 18e22 mo corrected age 201/3749 (5.4) 20/422 (4.7) 1.01 (0.65e1.58)c
ANS, antenatal corticosteroid; ARR, adjusted relative risk; CI, confidence interval; CPAP, continuous positive airway pressure; VEGF, vascular endothelial growth factor.
a
ARR and 95% CI are estimated with no ANS (not exposed to ANS) group used as referent category except for complete vs partial column where ARR and 95% CI are expressed for complete course of
ANS compared to partial course of ANSemodels adjust for birthweight, sex, multiple births, small for gestational age, maternal variables (age, marital status, race, diabetes, rupture of membranes

24 h, antepartum hemorrhage, and mode of delivery), center, and epoch; b Significant with P value <.05; c Model does not adjust for centeremodel did not converge with center included.
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018.

corticosteroids compared to infants without exposure (Table 2). The rate of dysplasia, and prolonged hospital stay
without exposure (Table 2). pulmonary hemorrhage, pneumo- did not differ by antenatal corticoste-
Infants exposed to any antenatal cor- thorax, early-onset sepsis, proven roids exposure (Table 2). Among survi-
ticosteroids had a lower rate of severe necrotizing enterocolitis, patent ductus vors exposed to any antenatal
intracranial hemorrhage/periventricular arteriosus treated with indomethacin/ corticosteroids there was a higher rate of
leukomalacia and severe retinopathy of ibuprofen, treatment with postnatal oxygen therapy at discharge compared to
prematurity compared to infants steroids for bronchopulmonary infants without exposure but among

JANUARY 2018 American Journal of Obstetrics & Gynecology 130.e4

Original Research OBSTETRICS ajog.org

survivors who were followed at 18-22

FIGURE 2
months’ corrected age, the rate of oxygen
Death before discharge by antenatal corticosteroid exposure
therapy and continuous positive airway
pressure/ventilator did not differ by
antenatal corticosteroid exposure
(Table 2).
Infants exposed to a complete course
of antenatal corticosteroids had a lower
rate of death before discharge
compared to infants exposed to a
partial course of antenatal corticoste-
roids (Supplementary Table 1). Infants
exposed to a complete course of
antenatal corticosteroids also had a
lower rate of death due to respiratory
distress syndrome, surfactant use,
mechanical ventilation, severe intra-
cranial hemorrhage/periventricular
leukomalacia, and pulmonary hemor-
Death before discharge of infants 22-28 weeks’ gestation by antenatal corticosteroid (ANS) expo-
rhage compared to infants exposed to a sure. Infants exposed to course of ANS had lower rate of death before discharge compared to infants
partial course of antenatal corticoste- who were not exposed to ANS at each gestational age from 23-27 weeks. Infants exposed to partial
roids (Supplementary Table 1). course of ANS had lower mortality at 23, 24, and 26 weeks’ gestation.
The inclusion of the 883 infants who Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018.
died within 12 hours without receiving
delivery room resuscitation strength-
ened the association between lower
mortality and exposure to antenatal and mechanical ventilation among used were different.21,22 In the Cochrane
corticosteroids, particularly among infants exposed to antenatal corticoste- review subgroup analysis of those infants
infants at the lowest gestations roids indicating amelioration of the <28 weeks’ gestation, there was no
(Supplementary Table 2). The results of course of respiratory distress syndrome. significant reduction in the rate of res-
the models using multiple imputation piratory distress syndrome (4 studies,
for missing data were substantively Meaning/clinical implications of 102 infants).1 Rates of bronchopulmo-
similar to the primary analysis findings nary dysplasia were not analyzed by
(Supplementary Table 3). The results of The current study demonstrates a lower gestational age at delivery but were not
the models that included chorioamnio- rate of death associated with exposure to different for the cohort as a whole (6
nitis did not substantively differ and are antenatal corticosteroids in infants <29 studies, 818 infants). In the current
not presented. weeks’ gestation. Randomized study, the rates of respiratory distress
controlled trials of administration of syndrome and the rates of broncho-
Comment antenatal corticosteroids to women who pulmonary dysplasia among survivors
Principal findings delivered at <30 weeks’ gestation show were also not different.
This large multicenter observational inconclusive neonatal outcomes, in part Observational studies have shown
study shows that exposure to a complete due to small sample size.1,2 In the marked variability in pulmonary out-
or partial course of antenatal corticoste- Cochrane review subgroup analysis of comes in infants at the lowest gestations.
roids is associated with lower mortality those infants <28 weeks’ gestation,1 A NICHD Neonatal Research Network
in infants 22-28 weeks’ gestation and there was no significant reduction in study including 10,541 infants delivered
weighing
401 g after adjustment for the rate of neonatal death but the sample between 22-25 weeks’ gestation found
multiple confounders. Among survivors, was small (2 studies, 89 infants). The that a lower mortality rate associated
the rate of bronchopulmonary dysplasia current study agrees with data from with exposure to antenatal corticoste-
in infants exposed to either a complete or observational studies that show a sur- roids was partially offset by a higher rate
partial course of antenatal corticoste- vival benefit among extremely preterm of bronchopulmonary dysplasia among
roids did not differ compared to infants infants at the lowest gestations.20 survivors compared to infants without
who were not exposed. Although all Many of the studies included in the exposure.3 A multicenter study of 2549
groups had a high rate of respiratory meta-analyses of randomized controlled infants <29 weeks’ gestation also found
distress syndrome, there was a lower rate trials1,2 were carried out in the era before that the rate of bronchopulmonary
of death due to respiratory distress the widespread use of surfactant and dysplasia was higher in infants exposed
syndrome and a lower use of surfactant when the methods of ventilatory support to a complete course of antenatal

130.e5 American Journal of Obstetrics & Gynecology JANUARY 2018

ajog.org OBSTETRICS Original Research

TABLE 3
Selected outcomes of infants 22e28 weeks’ gestation by antenatal corticosteroid treatment for births from 2006
through 2014
Complete ANS Partial ANS No ANS
n/total n (%) ARR (95% CI)a n/total n (%) ARR (95% CI) n/total n (%)
Bronchopulmonary
dysplasia, physiologic, or death
by 36 wk’ postmenstrual age
22 wk 45/48 (93.8) 0.95 (0.86e1.05) 27/28 (96.4) 0.96 (0.88e1.05) 94/96 (97.9)
b
23 wk 437/508 (86.0) 0.89 (0.84e0.94) 282/312 (90.4) 0.98 (0.93e1.02) 264/281 (94.0)
24 wk 1100/1390 (79.1) 0.95 (0.87e1.03) 404/499 (81.0) 1.03 (0.94e1.12) 166/208 (79.8)
25 wk 1043/1581 (66.0) 0.93 (0.84e1.02)c 374/572 (65.4) 0.94 (0.84e1.05) 163/250 (65.2)
c
26 wk 835/1483 (56.3) 0.93 (0.83e1.04) 283/495 (57.2) 0.97 (0.84e1.11) 144/245 (58.8)
27 wk 592/1243 (47.6) 0.86 (0.72e1.02)c 159/367 (43.3) 0.84 (0.69e1.03)c 74/132 (56.1)
c c
28 wk 330/781 (42.3) 0.98 (0.75e1.27) 85/231 (36.8) 0.88 (0.64e1.20) 35/88 (39.8)
Death by 36 wk’
postmenstrual age
22 wk 31/48 (64.6) 0.79 (0.61e1.04)c 19/29 (65.5) 0.85 (0.64e1.14)c 78/97 (80.4)
b,c b,c
23 wk 226/509 (44.4) 0.68 (0.59e0.78) 156/314 (49.7) 0.77 (0.67e0.88) 194/281 (69.0)
b,c b
24 wk 441/1409 (31.3) 0.54 (0.45e0.64) 170/504 (33.7) 0.75 (0.60e0.93) 99/209 (47.4)
25 wk 310/1596 (19.4) 0.66 (0.51e0.85)b,c 122/576 (21.2) 0.76 (0.58e0.99)b,c 68/253 (26.9)
b,c b,c
26 wk 176/1499 (11.7) 0.55 (0.40e0.76) 69/499 (13.8) 0.64 (0.45e0.92) 47/247 (19.0)
b,c c
27 wk 112/1262 (8.9) 0.60 (0.37e0.98) 37/372 (9.9) 0.73 (0.42e1.27) 20/130 (15.4)
c c
28 wk 56/796 (7.0) 0.82 (0.39e1.74) 18/233 (7.7) 0.75 (0.30e1.87) 7/88 (8.0)
Death before discharge
22 wk 32/48 (66.7) 0.83 (0.64e1.07)c 19/29 (65.5) 0.83 (0.63e1.10)c 79/96 (82.3)
b,c b,c
23 wk 243/505 (48.1) 0.72 (0.63e0.82) 164/310 (52.9) 0.80 (0.70e0.91) 199/280 (71.1)
b,c b
24 wk 480/1406 (34.1) 0.57 (0.48e0.67) 184/504 (36.5) 0.75 (0.61e0.91) 105/207 (50.7)
25 wk 352/1592 (22.1) 0.75 (0.59e0.95)b,c 136/574 (23.7) 0.83 (0.64e1.07)c 71/253 (28.1)
b,c b,c
26 wk 210/1498 (14.0) 0.61 (0.45e0.82) 78/499 (15.6) 0.67 (0.48e0.94) 51/247 (20.6)
b,c c
27 wk 138/1260 (11.0) 0.58 (0.38e0.87) 47/371 (12.7) 0.70 (0.44e1.13) 26/131 (19.8)
28 wk 74/795 (9.3) 0.86 (0.44e1.67)c 26/233 (11.2) 1.06 (0.49e2.27)c 9/88 (10.2)
Bronchopulmonary dysplasia,
physiologic definition
22 wk 13/16 (81.3) 0.64 (0.35e1.16) 8/9 (88.9) 0.96 (0.82e1.11)c 15/16 (93.8)
b,c
23 wk 191/261 (73.2) 0.80 (0.68e0.94) 114/144 (79.2) 0.97 (0.83e1.12)c 64/81 (79.0)
c
24 wk 619/907 (68.2) 1.09 (0.92e1.30) 220/315 (69.8) 1.18 (0.98e1.41) 59/101 (58.4)
c
25 wk 690/1224 (56.4) 0.96 (0.83e1.12) 235/434 (54.1) 0.95 (0.81e1.12) 92/179 (51.4)
26 wk 622/1268 (49.1) 0.99 (0.85e1.15)c 206/417 (49.4) 0.95 (0.80e1.12)c 93/194 (47.9)
c c
27 wk 454/1103 (41.2) 0.94 (0.74e1.19) 111/319 (34.8) 0.87 (0.66e1.15) 47/105 (44.8)
c c
28 wk 258/706 (36.5) 1.02 (0.74e1.42) 59/205 (28.8) 0.83 (0.56e1.22) 26/79 (32.9)
Respiratory support at
discharge, oxygen
d
22 wk 13/16 (81.3) 6/10 (60.0) 0.52 (0.23e1.18)c 9/14 (64.3)
c d
23 wk 150/258 (58.1) 1.04 (0.78e1.39) 82/140 (58.6) 36/77 (46.8)
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018. (continued)

JANUARY 2018 American Journal of Obstetrics & Gynecology 130.e6

Original Research OBSTETRICS ajog.org

TABLE 3
Selected outcomes of infants 22e28 weeks’ gestation by antenatal corticosteroid treatment for births from 2006
through 2014 (continued)
Complete ANS Partial ANS No ANS
a
n/total n (%) ARR (95% CI) n/total n (%) ARR (95% CI) n/total n (%)
24 wk 407/892 (45.6) 1.26 (0.94e1.69)c 134/309 (43.4) 1.17 (0.86e1.60)c 34/101 (33.7)
c c
25 wk 434/1196 (36.3) 1.06 (0.83e1.36) 158/431 (36.7) 1.22 (0.92e1.60) 50/179 (27.9)
c c
26 wk 352/1238 (28.4) 1.21 (0.90e1.62) 124/409 (30.3) 1.23 (0.90e1.68) 44/191 (23.0)
27 wk 278/1077 (25.8) 1.39 (0.90e2.13)c 63/317 (19.9) 1.22 (0.74e2.03)c 18/101 (17.8)
c d
28 wk 144/688 (20.9) 3.62 (1.38e9.46) 38/202 (18.8) 4/77 (5.2)
ANS, antenatal corticosteroid; ARR, adjusted relative risk; CI, confidence interval.
a
ARR and 95% CI are estimated with no ANS (not exposed to ANS) group used as referent categoryemodels adjust for birthweight, sex, multiple births, small for gestational age, maternal variables
(age, marital status, race, diabetes, rupture of membranes
24 h, antepartum hemorrhage, and mode of delivery), center, and epoch; b Significant with P value <.05; c Model does not adjust for
centeremodel did not converge with center included; d Model did not converge even with center excludedefor respiratory support at discharge (oxygen): unadjusted relative risks (95% CI) for
complete ANS at 22 wk: 1.26 (0.80e1.99); for partial ANS at 23 wk: 1.25 (0.95e1.65); and for partial ANS at 28 wk: 3.62 (1.34e9.81).
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018.

corticosteroids compared to infants g from the era before widespread use of confounding, there may be some residual
without exposure.4 Another multicenter antenatal corticosteroids and surfactant unmeasured bias in the results due to
study of 11,607 infants born 22-33 that found a lower rate of death before baseline differences between the study
weeks’ gestation found that the rates of discharge in infants exposed to a partial groups that may not be adequately
bronchopulmonary dysplasia did not course of antenatal corticosteroids adjusted for in the models used. In
differ in infants exposed to any antenatal compared to infants without exposure.6 addition, there is a possibility of postnatal
corticosteroids compared to infants In that study the rate of broncho- bias in which infants not exposed to
without exposure.5 These differences in pulmonary dysplasia also did not differ antenatal corticosteroids may have had
bronchopulmonary dysplasia rates may by degree of exposure to antenatal cor- their care restricted or withheld, partic-
be related to the different definition of ticosteroids. The aforementioned study ularly among infants at the lowest gesta-
bronchopulmonary dysplasia9 used in by Wong et al4 indicated that mortality tions. To reduce this effect, infants who
these studies as well as the gestational age did not differ in infants exposed to a died within 12 hours of birth without
inclusion criteria as inclusion of infants partial course of antenatal corticoste- receiving delivery room resuscitation
at the lowest gestations would result in roids compared to infants without were excluded from the primary analysis.
more survivors who can develop bron- exposure but this study was limited by a The multiple testing used in this study at a
chopulmonary dysplasia. relatively small sample size. 5% significance level may have resulted in
An important focus of the current a few results being significant purely by
study was the differential benefits of a Strengths and weaknesses chance. However, the benefits were
partial or a complete course of antenatal This study used data collected from top consistent at most gestations, suggesting
corticosteroids. Although a complete academic centers across the United States that the results are not only due to chance.
course of antenatal corticosteroids is where optimal obstetric and neonatal
associated with a lower mortality care might be anticipated but there are Research implications
compared to a partial course, the current several limitations that should be noted. Although antenatal corticosteroid
study indicates that the first dose of There was no inception cohort of fetuses administration reduces preterm infant
antenatal corticosteroids may have the exposed or not exposed to antenatal mortality and morbidity without
largest effect on reducing mortality. The corticosteroids. Data were not available increasing the cost of care,1 many eligible
Cochrane review of randomized on the exact timing of antenatal cortico- women24 do not receive this treat-
controlled trials subgroup analysis of steroids, whether fetal monitoring was ment.11,12 Center differences in the use
infants delivered following a partial undertaken prior to delivery, or the of antenatal corticosteroids are associ-
course of antenatal corticosteroids length of maternal hospitalization before ated with mortality among infants at the
showed a significant reduction in delivery.23 There is a risk of bias as lowest gestations.25,26 Differences
neonatal death in infants exposed to a women admitted in advanced labor most between administration rates among
partial course (4 studies, 295 infants).1 likely would be overrepresented in the infants by gestation in this study indicate
The current study results are in concor- group that did not receive antenatal cor- that although administration rates are
dance with those of an observational ticosteroids. While it is also unlikely that increasing, there are opportunities for
study of 9949 infants weighing 501-1500 the results of this study are only due to quality improvement.26

130.e7 American Journal of Obstetrics & Gynecology JANUARY 2018

ajog.org OBSTETRICS Original Research

Conclusion Duke University School of Medicine, Univer- Brudos, PhD; Alexis S. Davis, MD, MS; Maria
The current study demonstrates that sity Hospital, Duke Regional Hospital, and Elena DeAnda, PhD; Anne M. DeBattista, RN,
University of North Carolina (U10 HD40492, PNP; Barry E. Fleisher, MD; Lynne C. Huffman,
antenatal exposure to corticosteroids for M01 RR30)eC. Michael Cotten, MD, MHS; Ricki MD; Jean G. Kohn, MD, MPH; Casey Krueger,
infants 22 0/7 to 28 6/7 weeks’ gestation F. Goldstein, MD; Kathy J. Auten, MSHS; PhD; Julie C. Lee-Ancajas, PhD; Andrew W.
is associated with a lower rate of death Joanne Finkle, RN, JD; Kimberley A. Fisher, Palmquist, RN; Melinda S. Proud, RCP; Renee
before discharge without a higher rate of PhD, FNP-BC, IBCLC; Katherine A. Foy, RN; P. Pyle, PhD; Dharshi Sivakumar, MD; Robert D.
bronchopulmonary dysplasia or other Sandra Grimes, RN, BSN; Kathryn E. Gus- Stebbins, MD; Nicholas H. St John, PhD; Halie
tafson, PhD; Melody B. Lohmeyer, RN, MSN; E. Weiss, MD.
major adverse pulmonary problems. Matthew M. Laughon, MD, MPH; Carl L. Bose, Tufts Medical Center, Floating Hospital for
This study also indicates that antenatal MD; Janice Bernhardt, MS, RN; Gennie Bose, Children (U10 HD53119, M01 RR54)eIvan D.
corticosteroids ameliorate the severity of RN; Cynthia L. Clark, RN. Frantz III, MD; Elisabeth C. McGowan, MD;
respiratory distress syndrome and other Emory University, Children’s Healthcare of Brenda L. MacKinnon, RNC; Ellen Nylen, RN,
important morbidities in extremely Atlanta, Grady Memorial Hospital, and Emory BSN; Anne Furey, MPH; Ana Brussa, MS, OTR/
University Hospital Midtown (U10 HD27851, L; Cecelia Sibley, PT, MHA.
preterm infants. n M01 RR39)eDavid P. Carlton, MD; Ira Adams- University of Alabama at Birmingham Health
Chapman, MD; Yvonne C. Loggins, RN, BSN; System and Children’s Hospital of Alabama
Diane I. Bottcher, RN, MSN; Maureen Mulligan (U10 HD34216, M01 RR32)eMyriam Peralta-
Acknowledgment LaRossa, RN; Sheena L. Carter, PhD. Carcelen, MD, MPH; Kathleen G. Nelson, MD;
We are indebted to our medical and nursing Eunice Kennedy Shriver National Institute of Kirstin J. Bailey, PhD; Fred J. Biasini, PhD; Ste-
colleagues and the infants and their parents who Child Health and Human DevelopmenteLinda L. phanie A. Chopko, PhD; Monica V. Collins, RN,
agreed to take part in this study. The following Wright, MD; Elizabeth M. McClure, MEd; Ste- BSN, MaEd; Shirley S. Cosby, RN, BSN; Mary
investigators, in addition to those listed as au- phanie Wilson Archer, MA. Beth Moses, PT, MS, PCS; Vivien A. Phillips, RN,
thors, participated in this study: Indiana University, University Hospital, Meth- BSN; Julie Preskitt, MSOT, MPH; Richard V.
Neonatal Research Network Steering Com- odist Hospital, Riley Hospital for Children, and Rector, PhD; Sally Whitley, MA, OTR-L, FAOTA.
mittee Chairs: Alan H. Jobe, MD, PhD, University Wishard Health Services (U10 HD27856, M01 University of CaliforniaeLos Angeles, Mattel
of Cincinnati (2003 through 2006); Michael S. RR750)eGregory M. Sokol, MD; Brenda B. Children’s Hospital, Santa Monica Hospital, Los
Caplan, MD, University of Chicago, Pritzker Poindexter, MD, MS; James A. Lemons, MD; Robles Hospital and Medical Center, and Olive
School of Medicine (2006 through 2011); and Anna M. Dusick, MD; Carolyn Lytle, MD, MPH; View Medical Center (U10 HD68270)eUday
Richard A. Polin, MD, Division of Neonatology, Lon G. Bohnke, MS; Greg Eaken, PhD; Faithe Devaskar, MD; Meena Garg, MD; Isabell B.
College of Physicians and Surgeons, Columbia Hamer, BS; Dianne E. Herron, RN; Abbey Hines, Purdy, PhD, CPNP; Teresa Chanlaw, MPH;
University (2011 through present). PsyD; Lucy C. Miller, RN, BSN, CCRC; Heike M. Rachel Geller, RN, BSN.
Alpert Medical School of Brown University Minnich, PsyD, HSPP; Lu-Ann Papile, MD; Leslie University of CaliforniaeSan Diego Medical
and Women & Infants Hospital of Rhode Island Richard, RN; Leslie Dawn Wilson, BSN CCRC. Center and Sharp Mary Birch Hospital for
(U10 HD27904)eWilliam Oh, MD; Martin Kes- Nationwide Children’s Hospital and Ohio Women and Newborns (U10 HD40461)eNeil N.
zler, MD; Betty R. Vohr, MD; Robert T. Burke, State University Wexner Medical Center (U10 Finer, MD; Maynard R. Rasmussen, MD; Yvonne
MD, MPH; Bonnie E. Stephens, MD; Yvette HD68278)ePablo J. Sánchez, MD; Leif D. Nelin, E. Vaucher, MD, MPH; Paul R. Wozniak, MD;
Yatchmink, MD; Barbara Alksninis, RNC, PNP; MD; Sudarshan R. Jadcherla, MD; Patricia Kathy Arnell, RNC; Renee Bridge, RN; Clarence
Angelita M. Hensman, MS, RNC-NIC; Kristin Luzader, RN; Christine A. Fortney, RN, PhD; Demetrio, RN; Martha G. Fuller, RN, MSN; Wade
Basso, RN, MaT; Elisa Vieira, RN, BSN; Lenore Keith Yeates, PhD; Melanie Stein, BBA, RRT; Rich, BSHS, RRT.
Keszler, MD; Teresa M. Leach, MEd, CAES; Julie Gutentag, RN, BSN; Tiffany Sharp, CMDA; University of Iowa and Mercy Medical Center
Martha R. Leonard, BA, BS; Lucy Noel; Rachel Courtney Cira, RRT; Lina Yossef-Salameh, MD; (U10 HD53109, M01 RR59)eJohn A. Widness,
A. Vogt, MD; Victoria E. Watson, MS, CAS. Pamela Morehead, BS; Cody Brennan; Rox Ann MD; Dan L. Ellsbury, MD; Tarah T. Colaizy, MD,
Case Western Reserve University, Rainbow Sullivan, RN, BSN; Erin Fearns; Aubry Folwer; MPH; Michael J. Acarregui, MD; Jane E. Brum-
Babies & Children’s Hospital (U10 HD21364, Jennifer Notestine, RN; Cole Hague, BA, MS; baugh, MD; Karen J. Johnson, RN, BSN; Donia
M01 RR80)eAvroy A. Fanaroff, MD; Deanne E. Jennifer L. Grothause, RN; Bronte Clifford, BA; B. Campbell, RNC-NIC; Diane L. Eastman, RN,
Wilson-Costello, MD; Bonnie S. Siner, RN; Amanda Daubenmire Morely, BS; Erin Wishloff, CPNP, MA.
Harriet G. Friedman, MA. RRT; Sarah Keim, BA, MA, MS, PhD; Helen University of Miami, Holtz Children’s Hospital
Children’s Mercy Hospital, University of Carey, DHSC; Christopher Timan, MD. (U10 HD21397, M01 RR16587)eShahnaz
MissourieKansas City School of Medicine (U10 RTI International (U10 HD36790)eW. Ken- Duara, MD; Charles R. Bauer, MD; Ruth Everett-
HD68284)eWilliam E. Truog, MD; Eugenia K. neth Poole, PhD (deceased); Dennis Wallace, Thomas, RN, MSN; Amy Mur Worth, RN, MS;
Pallotto, MD, MSCE; Howard W. Kilbride, MD; PhD; Jamie E. Newman, PhD, MPH; Jeanette Mary Allison, RN; Alexis N. Diaz, BA; Elaine E.
Cheri Gauldin, RN, BS, CCRC; Anne Holmes, O’Donnell Auman, BS; Margaret M. Crawford, Mathews, RN; Kasey Hamlin-Smith, PhD; Lissa
RN, MSN, MBA-HCM, CCRC; Kathy Johnson, BS, CCRP; Betty K. Hastings; Elizabeth M. Jean-Gilles, BA; Maria Calejo, MS; Silvia M.
RN, CCRC; Allison Knutson, RN, BSN. McClure, MEd; Carolyn M. Petrie Huitema, MS, Frade Eguaras, BA; Silvia Fajardo-Hiriart, MD;
Cincinnati Children’s Hospital Medical Cen- CCRP; Kristin M. Zaterka-Baxter, RN, BSN, Yamiley C. Gideon, BA; Michelle Harwood Ber-
ter, University Hospital, and Good Samaritan CCRP. We thank Lei Li, PhD, for his assistance kovits, PhD; Alexandra Stoerger, BA; Andrea
Hospital (U10 HD27853, M01 RR8084)eKurt with multiple imputation analyses. Garcia, MA; Helena Pierre, BA; Georgette
Schibler, MD; Edward F. Donovan, MD; Kate Stanford University, California Pacific Medical Roder, BSW; Arielle Riguad, MD.
Bridges, MD; Jean J. Steichen, MD; Kimberly Center, Dominican Hospital, El Camino Hospital, University of New Mexico Health Sciences
Yolton, PhD; Barbara Alexander, RN; Estelle E. and Lucile Packard Children’s Hospital (U10 Center (U10 HD27881, U10 HD53089, M01
Fischer, MHSA, MBA; Cathy Grisby, BSN, HD27880, M01 RR70)eKrisa P. Van Meurs, RR997)eKristi L. Watterberg, MD; Andrea
CCRC; Marcia Worley Mersmann, RN; Holly L. MD; David K. Stevenson, MD; Susan R. Hintz, Freeman Duncan, MD, MScr; Janell Fuller, MD;
Mincey, RN, BSN; Jody Hessling, RN; Teresa L. MD, MS, Epi; Marian M. Adams, MD; Charles E. Robin K. Ohls, MD; Lu-Ann Papile, MD; Conra
Gratton, PA; Lenora D. Jackson, CRC; Kristin Ahlfors, MD; Joan M. Baran, PhD; Barbara Backstrom Lacy, RN; Sandra Brown, RN, BSN;
Kirker, CRC; Greg Muthig, BS. Bentley, PhD; Lori E. Bond, PhD; Ginger K. Jean R. Lowe, PhD; Rebecca Montman, BSN.

JANUARY 2018 American Journal of Obstetrics & Gynecology 130.e8

Original Research OBSTETRICS ajog.org

University of Pennsylvania, Hospital of the Shawna Rodgers, RN; Saba Khan Siddiki, MD; outcomes of very premature infants: a regional
University of Pennsylvania, Pennsylvania Hos- Maegan C. Simmons, RN; Daniel Sperry, RN; cohort study. Arch Dis Child Fetal Neonatal Ed
pital, and Children’s Hospital of Philadelphia Laura L. Whitely, MD; Sharon L. Wright, MT. 2014;99:F12-20.
(U10 HD68244)eHaresh Kirpalani, MB, MSc; University of Utah Medical Center, Inter- 5. Mori R, Kusuda S, Fujimura M; Neonatal
Sara B. DeMauro, MD, MSCE; Aasma S. mountain Medical Center, LDS Hospital, and Research Network. Antenatal corticosteroids
Chaudhary, BS, RRT; Soraya Abbasi, MD; Toni Primary Children’s Medical Center (U10 promote survival of extremely preterm infants
Mancini, RN, BSN, CCRC; Dara M. Cucinotta, HD53124, M01 RR64)eRoger G. Faix, MD; born at 22 to 23 weeks of gestation. J Pediatr
RN; Judy C. Bernbaum, MD; Marsha Gerdes, Bradley A. Yoder, MD; Michael Steffen, MS, 2011;159:110-4.
PhD; Hallam Hurt, MD. CPM; Shawna Baker, RN; Karie Bird, RN; Jill 6. Wright LL, Verter J, Younes N, et al.
University of Rochester Medical Center, Burnett, RN; Jennifer J. Jensen, RN, BSN; Antenatal corticosteroid administration and
Golisano Children’s Hospital, and University of Karen A. Osborne, RN, BSN, CCRC; Cynthia neonatal outcome in very low birth weight
Buffalo Women’s and Children’s Hospital of Spencer, RNC; Kimberlee Weaver-Lewis, RN, infants: the NICHD Neonatal Research
Buffalo (U10 HD68263, U10 HD40521, M01 MS; Sarah Winter, MD; Karen Zanetti, RN. Network. Am J Obstet Gynecol 1995;173:
RR44, UL1 TR42)eDale L. Phelps, MD; Ronnie Wake Forest University Baptist Medical 269-74.
Guillet, MD, PhD; Satyan Lakshminrusimha, MD; Center, Brenner Children’s Hospital, and For- 7. Manktelow BN, Lal MK, Field DJ, Sinha SK.
Gary J. Myers, MD; Erica Burnell, RN; Stephanie syth Medical Center (U10 HD40498, M01 Antenatal corticosteroids and neonatal out-
Guilford, BS; Cassandra A. Horihan, MS; Diane RR7122)eT. Michael O’Shea, MD, MPH; Robert comes according to gestational age: a cohort
Hust, MS, RN, CS; Rosemary L. Jensen; Julie G. Dillard, MD; Nancy J. Peters, RN, CCRP; study. Arch Dis Child Fetal Neonatal Ed
Babish Johnson, MSW; Emily Kushner, MA; Korinne Chiu, MA; Deborah Evans Allred, MA, 2010;95:F95-8.
Deanna Maffett, RN; Joan Merzbach, LMSW; LPA; Donald J. Goldstein, PhD; Raquel Halfond, 8. Doyle LW, Kitchen WH, Ford GW,
Linda J. Reubens, RN, CCRC; Ann Marie MA; Barbara G. Jackson, RN, BSN; Carroll Rickards AL, Kelly EA. Antenatal steroid therapy
Reynolds, MD; Mary Rowan, RN; Holly I. M. Peterson, MA; Dia Roberts, BSN, PNP; Ellen L. and 5-year outcome of extremely low birth
Wadkins, MA; Ashley Williams, MS Ed; Karen Waldrep, MS; Melissa Whalen Morris, MA; Gail weight infants. Obstet Gynecol 1989;73:743-6.
Wynn, NNP, RN; Kelly Yost, PhD; Lauren Wiley Hounshell, PhD; Lisa K. Washburn, MD; 9. Jobe AH, Bancalari E. Bronchopulmonary
Zwetsch, RN, MS, PNP. Cherrie D. Welch, MD, MPH. dysplasia. Am J Respir Crit Care Med 2001;163:
University of Tennessee Health Science Wayne State University, Hutzel Women’s 1723-9.
Center (U10 HD21415)eSheldon B. Korones, Hospital, and Children’s Hospital of Michigan 10. Brownfoot FC, Gagliardi DI, Bain E,
MD; Henrietta S. Bada, MD; Tina Hudson, RN, (U10 HD21385)eBeena G. Sood, MD, MS; Middleton P, Crowther CA. Different cortico-
BSN; Marilyn Williams, LCSW; Kimberly Yolton, Athina Pappas, MD; John Barks, MD; Yvette R. steroids and regimens for accelerating fetal lung
PhD. Johnson, MD, MPH; Rebecca Bara, RN, BSN; maturation for women at risk of preterm birth.
University of Texas Southwestern Medical Laura Goldston, MA; Mary E. Johnson, RN, Cochrane Database Syst Rev 2013;8:
Center, Parkland Health and Hospital System BSN; Deborah Kennedy, RN, BSN; Geraldine CD006764.
and Children’s Medical Center Dallas (U10 Muran, RN, BSN; Elizabeth Billian, RN, MBA; 11. Davidson C, Monga M, Ellison D, Vidaeff A.
HD40689, M01 RR633)eLuc P. Brion, MD; Laura Sumner, RN, BSN; Kara Sawaya, RN, Continuation of pregnancy after antenatal corti-
Pablo J. Sánchez, MD; Abbot R. Laptook, MD; BSN; Mary K. Christensen, BA, RRT; Stephanie costeroid administration: opportunity for
Charles R. Rosenfeld, MD; Walid A. Salhab, MD; A. Wiggins, MS. rescue? J Reprod Med 2010;55:14-8.
R. Sue Broyles, MD; Roy J. Heyne, MD; Merle Yale University, Yale-New Haven Children’s 12. Melamed N, Shah J, Soraisham A, et al.
Ipson, MD; Sally S. Adams, MS, RN, CPNP; P. Hospital, and Bridgeport Hospital (U10 Association between antenatal corticosteroid
Jeannette Burchfield, RN, BSN; Lijun Chen, HD27871, UL1 TR142, M01 RR125, M01 administration-to-birth interval and outcomes of
PhD, RN; Cristin Dooley, PhD, LSSP; Alicia RR6022)eRichard A. Ehrenkranz, MD; Christine preterm neonates. Obstet Gynecol 2015;125:
Guzman; Gaynelle Hensley, RN; Elizabeth T. Butler, MD; Harris Jacobs, MD; Patricia Cer- 1377-84.
Heyne, MS, MA, PA-C, PsyD; Jackie F. vone, RN; Nancy Close, PhD; Patricia Gettner, 13. Patel RM, Kandefer S, Walsh MC, et al;
Hickman, RN; Melissa H. Leps, RN; Linda A. RN; Walter Gilliam, PhD; Sheila Greisman, RN; for the Eunice Kennedy Shriver National Institute
Madden, BSN, RN, CPNP; Susie Madison, RN; Monica Konstantino, RN, BSN; JoAnn Poulsen, of Child Health and Human Development
Nancy A. Miller, RN; Janet S. Morgan, RN; RN; Elaine Romano, MSN; Janet Taft, RN, BSN; Neonatal Research Network. Causes and timing
Lizette E. Torres, RN; Cathy Twell Boatman, MS, Joanne Williams, RN, BSN. of death in extremely premature infants from
CIMI; Diana M. Vasil, RNC-NIC. 2000 through 2011. N Engl J Med 2015;372:
University of Texas Health Science Center at 331-40.
Houston Medical School, Children’s Memorial References 14. McKinlay CJ, Crowther CA, Middleton P,
Hermann Hospital, and Lyndon Baines Johnson 1. Roberts D, Brown J, Medley N, Dalziel SR. Harding JE. Repeat antenatal glucocorticoids
General Hospital/Harris County Hospital District Antenatal corticosteroids for accelerating fetal for women at risk of preterm birth: a Cochrane
(U10 HD21373)eKathleen A. Kennedy, MD, lung maturation for women at risk of preterm Systematic Review. Am J Obstet Gynecol
MPH; Jon E. Tyson, MD, MPH; Patrick M. birth. Cochrane Database Syst Rev 2017;3: 2012;206:187-94.
Jones, MD; Patricia W. Evans, MD; Esther G. CD004454. 15. Walsh MC, Wilson-Costello D, Zadell A,
Akpa, RN, BSN; Nora I. Alaniz, BS; Julie Arldt- 2. Onland W, de Laat MW, Mol BW, Offringa M. Newman N, Fanaroff A. Safety, reliability, and
McAlister, RN, BSN; Katrina Burson, RN, BSN; Effects of antenatal corticosteroids given prior to validity of a physiologic definition of broncho-
Magda Cedillo Guajardo, RB, BSN, FAACM; 26 weeks’ gestation: a systematic review of pulmonary dysplasia. J Perinatol 2003;23:
Susan E. Dieterich, PhD; Beverly Foley Harris, randomized controlled trials. Am J Perinatol 451-6.
RN, BSN; Claudia I. Franco, RNC, MSN; Car- 2011;28:33-44. 16. Stoll BJ, Hansen NI, Bell EF, et al; for the
men Garcia, RN, CCRP; Charles Green, PhD; 3. Carlo WA, McDonald SA, Fanaroff AA, et al. NICHD Neonatal Research Network. Trends in
Margarita Jiminez, MD, MPH; Janice John, for the NICHD Neonatal Research Network. care practices, morbidity, and mortality of
CPNP; Anna E. Lis, RN, BSN; Terri Major- Association of antenatal corticosteroids with extremely preterm neonates, 1993-2012. JAMA
Kincade, MD, MPH; Karen Martin, RN; Sara C. mortality and neurodevelopmental outcomes 2015;314:1039-51.
Martin, RN, BSN; Georgia E. McDavid, RN; among infants born at 22 to 25 weeks’ gesta- 17. Tyson JE, Parikh NA, Langer J, Green C,
Brenda H. Morris, MD; Patricia Ann Orekoya, tion. JAMA 2011;306:2348-58. Higgins RD. Intensive care for extreme
RN, BSN; Patti L. Pierce Tate, RCP; M. Layne 4. Wong D, Abdel-Latif M, Kent A; NICUS prematurityemoving beyond gestational age.
Poundstone, RN, BSN; Stacey Reddoch, BA; Network. Antenatal steroid exposure and N Engl J Med 2008;358:1672-81.

130.e9 American Journal of Obstetrics & Gynecology JANUARY 2018

ajog.org OBSTETRICS Original Research

18. Ambalavanan N, Van Meurs KP, Perritt R, Medicine Units. Obstetric determinants of
et al; for the NICHD Neonatal Research Network. neonatal survival: influence of willingness to Author and article information
Predictors of death or bronchopulmonary perform cesarean delivery on survival of From the Eunice Kennedy Shriver National Institute of
dysplasia in preterm infants with respiratory fail- extremely low-birth-weight infants. Am J Obstet Child Health and Human Development Neonatal Research
ure. J Perinatol 2008;28:420-6. Gynecol 1997;176:960-6. Network, Bethesda, MD (all authors), and the University of
19. Haas DM, Sischy AC, McCullough W, 24. Raju TN, Mercer BM, Burchfield DJ, Alabama at Birmingham, Birmingham, AL (Dr Carlo).
Simsiman AJ. Maternal ethnicity influences on Joseph GF Jr. Periviable birth: executive sum- Received Aug. 16, 2017; revised Oct. 23, 2017;
neonatal respiratory outcomes after antenatal mary of a joint workshop by the Eunice Kennedy accepted Nov. 6, 2017.
corticosteroid use for anticipated preterm de- Shriver National Institute of Child Health and The National Institutes of Health (NIH), the Eunice
livery. J Matern Fetal Neonatal Med 2011;24: Human Development, Society for Maternal-Fetal Kennedy Shriver National Institute of Child Health and
516-20. Medicine, American Academy of Pediatrics, and Human Development (NICHD), the National Center for
20. Park CK, Isayama T, McDonald SD. American College of Obstetricians and Gyne- Research Resources, and the National Center for
Antenatal corticosteroid therapy before 24 cologists. Am J Obstet Gynecol 2014;210: Advancing Translational Sciences provided grant support
weeks of gestation: a systematic review and 406-17. for the Neonatal Research Network Generic Database and
meta-analysis. Obstet Gynecol 2016;127: 25. Smith PB, Ambalavanan N, Li L, et al; for the follow-up studies through cooperative agreements. Dr
715-25. Eunice Kennedy Shriver National Institute of Travers is supported by Agency for Healthcare Research
21. Fischer HS, Bührer C. Avoiding endotra- Child Health and Human Development Neonatal and Quality (AHRQ) grant 5T32HS013852-14. While
cheal ventilation to prevent bronchopulmonary Research Network. Approach to infants born at NICHD staff did have input into the study design, conduct,
dysplasia: a meta-analysis. Pediatrics 22 to 24 weeks’ gestation: relationship to out- analysis, and manuscript drafting, the content is solely
2013;132:1351-60. comes of more-mature infants. Pediatrics the responsibility of the authors and does not necessarily
22. St John EB, Carlo WA. Respiratory distress 2012;129:1508-16. represent the official views of the NIH or AHRQ.
syndrome in VLBW infants: changes in man- 26. Rysavy MA, Li L, Bell EF, et al; for the Disclosure: Dr Carlo is on the board of MEDNAX Inc;
agement and outcomes observed by the NICHD Eunice Kennedy Shriver National Institute of there are no other relationships or activities that could
Neonatal Research Network. Semin Perinatol Child Health and Human Development appear to have influenced the submitted work.
2003;27:288-92. Neonatal Research Network. Between-hospi- Presented at the annual meeting of the Pediatric Ac-
23. Bottoms SF, Paul RH, Iams JD, et al. for the tal variation in treatment and outcomes in ademic Societies, San Diego, CA, April 25-28, 2015.
National Institute of Child Health and Human extremely preterm infants. N Engl J Med Corresponding author: Waldemar A. Carlo, MD.
Development Network of Maternal-Fetal 2015;372:1801-11. [email protected]

JANUARY 2018 American Journal of Obstetrics & Gynecology 130.e10

 Original Research
130.e11 American Journal of Obstetrics & Gynecology JANUARY 2018

SUPPLEMENTARY TABLE 1
Outcomes of infants by level of exposure to antenatal corticosteroids
Complete ANS Partial ANS No ANS Complete vs partial
a
n/total n (%) ARR (95% CI) n/total n (%) ARR (95% CI) n/total n (%) ARR (95% CI)
Total study population N ¼ 7136 N ¼ 2533 N ¼ 1307
Death
By 36 wk’ postmenstrual age 1352/7119 (19.0) 0.62 (0.57e0.68)b 591/2527 (23.4) 0.75 (0.68e0.83)b 513/1305 (39.3) 0.82 (0.76e0.90)b
Before discharge 1529/7104 (21.5) 0.66 (0.61e0.72)b 654/2520 (26.0) 0.77 (0.70e0.85)b 540/1302 (41.5) 0.85 (0.78e0.92)b
b,c c
1.52 (1.03e2.25)b,c

OBSTETRICS
Due to bronchopulmonary dysplasia 138/7098 (1.9) 1.88 (1.08e3.27) 33/2518 (1.3) 1.12 (0.59e2.11) 16/1299 (1.2)
Due to respiratory distress syndrome 475/7098 (6.7) 0.66 (0.55e0.80)b 222/2518 (8.8) 0.81 (0.66e0.99)b,c 171/1299 (13.2) 0.80 (0.68e0.94)b,c
Bronchopulmonary dysplasia [physiologic definition] or 4382/7034 (62.3) 0.92 (0.89e0.96)b 1614/2504 (64.5) 0.97 (0.92e1.01) 940/1300 (72.3) 0.97 (0.94e1.01)
death by 36 wk’ postmenstrual age, 2006 through 2014
Population of survivors N ¼ 5575 N ¼ 1866 N ¼ 762
Bronchopulmonary dysplasia, physiologic definition 2847/5485 (51.9) 0.95 (0.88e1.02) 953/1843 (51.7) 0.98 (0.90e1.06) 396/755 (52.5) 0.999 (0.95e1.05)
Bronchopulmonary dysplasia, by use of supplemental 2982/5545 (53.8) 0.94 (0.88e1.01) 1008/1853 (54.4) 0.98 (0.91e1.06) 415/759 (54.7) 0.98 (0.93e1.02)
oxygen at 36 wk’ postmenstrual age, clinical
Respiratory distress syndrome 5478/5575 (98.3) 0.99 (0.99e0.998)b 1853/1866 (99.3) 1.00 (0.99e1.01) 760/762 (99.7) 0.99 (0.98e0.99)b
Surfactant use 4651/5575 (83.4) 0.90 (0.87e0.92)b 1665/1866 (89.2) 0.97 (0.94e0.99)b 702/762 (92.1) 0.93 (0.92e0.95)b
Mechanical ventilation 4973/5571 (89.3) 0.94 (0.92e0.96)b 1737/1865 (93.4) 0.98 (0.96e1.00) 723/762 (94.9) 0.96 (0.94e0.97)b
c c
Pneumothorax 256/5575 (4.6) 0.74 (0.53e1.04) 98/1866 (5.3) 0.88 (0.60e1.27) 45/762 (5.9) 0.84 (0.66e1.07)c
Pulmonary hemorrhage 203/5575 (3.6) 0.68 (0.49e0.94)b,c 101/1866 (5.4) 0.95 (0.67e1.36)c 47/762 (6.2) 0.73 (0.57e0.93)b,c
Treatment with postnatal steroids for bronchopulmonary 893/5420 (16.5) 0.97 (0.80e1.17) 320/1813 (17.7) 1.05 (0.86e1.28) 112/733 (15.3) 0.88 (0.78e0.98)b,c
dysplasia
Early-onset sepsis 99/5575 (1.8) 0.61 (0.36e1.04)c 35/1866 (1.9) 0.82 (0.46e1.45)c 17/762 (2.2) 0.73 (0.49e1.10)c
Necrotizing enterocolitis stage
2 472/5575 (8.5) 0.96 (0.75e1.24)c 193/1865 (10.3) 1.10 (0.84e1.44)c 71/762 (9.3) 0.86 (0.72e1.03)c
Intracranial hemorrhage/periventricular leukomalacia 664/5549 (12.0) 0.56 (0.48e0.66)b,c 360/1860 (19.4) 0.88 (0.74e1.04)c 167/761 (21.9) 0.63 (0.56e0.71)b,c
Patent ductus arteriosus treated with indomethacin/ 1743/5570 (31.3) 0.90 (0.81e0.99)b 697/1865 (37.4) 1.04 (0.93e1.16) 293/761 (38.5) 0.90 (0.84e0.96)b
ibuprofen
Retinopathy of prematurity stage
3 or treated with 932/5483 (17.0) 0.71 (0.61e0.82)b,c 388/1827 (21.2) 0.87 (0.74e1.02)c 178/753 (23.6) 0.82 (0.74e0.91)b,c
ablation/anti-VEGF drug
Respiratory support at discharge, oxygen 1778/5365 (33.1) 1.17 (1.03e1.33)b,c 605/1818 (33.3) 1.20 (1.05e1.38)b,c 195/740 (26.4) 0.98 (0.90e1.06)c
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018. (continued)

ajog.org
 ajog.org
SUPPLEMENTARY TABLE 1
Outcomes of infants by level of exposure to antenatal corticosteroids (continued)
Complete ANS Partial ANS No ANS Complete vs partial
a
n/total n (%) ARR (95% CI) n/total n (%) ARR (95% CI) n/total n (%) ARR (95% CI)
Prolonged hospital stay
120 d, all causes 1526/5442 (28.0) 0.86 (0.77e0.95)b,c 554/1839 (30.1) 1.00 (0.88e1.13) 235/752 (31.3) 0.92 (0.85e0.99)b
Population of survivors eligible for follow-up N ¼ 2981 N ¼ 1145 N ¼ 471
c d
Respiratory support at 18e22 mo corrected age, 49/2682 (1.8) 0.94 (0.41e2.14) 18/1045 (1.7) 1.11 (0.46e2.69) 7/422 (1.7) 0.93 (0.51e1.69)c
ventilation/CPAP
Oxygen at 18e22 mo corrected age 141/2682 (5.3) 0.99 (0.62e1.57)c 60/1045 (5.7) 1.24 (0.77e2.00)c 20/422 (4.7) 0.80 (0.58e1.11)c,e
ANS, antenatal corticosteroid; ARR, adjusted relative risk; CI, confidence interval; CPAP, continuous positive airway pressure; VEGF, vascular endothelial growth factor.
a
ARR and 95% CI are estimated with no ANS (not exposed to ANS) group used as referent category except for complete vs partial column where adjusted odds ratios and 95% CI are expressed for complete course of ANS compared to partial course of ANSemodels
adjust for birthweight, sex, multiple births, small for gestational age, maternal variables (age, marital status, race, diabetes, rupture of membranes
24 h, antepartum hemorrhage, and mode of delivery), center, and epoch; b Significant with P value <.05; c Does
not adjust for centeremodel did not converge with center included; d Model does not adjust for center or diabetesemodel did not converge with either variable included; e Significant interaction from separate models that also included ANS-epoch.
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018.
JANUARY 2018 American Journal of Obstetrics & Gynecology

OBSTETRICS
Original Research
130.e12
Original Research OBSTETRICS ajog.org

SUPPLEMENTARY TABLE 2
Death before discharge of infants 22e28 weeks’ gestation by antenatal corticosteroid treatment
Any ANS No ANS
N ¼ 9827 N ¼ 2028
Gestational age n/total n (%) n/total n (%) ARR (95% CI)a
22 wk 80/106 (75.5) 519/536 (96.8) 0.79 (0.70e0.90)b,c
23 wk 494/904 (54.6) 438/519 (84.4) 0.68 (0.63e0.74)b,c
24 wk 696/1947 (35.7) 134/236 (56.8) 0.58 (0.50e0.66)b,c
25 wk 496/2183 (22.7) 78/260 (30.0) 0.78 (0.61e1.00)
26 wk 294/2011 (14.6) 55/251 (21.9) 0.60 (0.45e0.79)b,c
27 wk 189/1644 (11.5) 29/134 (21.6) 0.59 (0.40e0.89)b,c
28 wk 101/1032 (9.8) 13/92 (14.1) 0.69 (0.39e1.22)c
Includes data on 883 infants who died within 12 h without receiving delivery room resuscitation.
ANS, antenatal corticosteroid; ARR, adjusted relative risk; CI, confidence interval.
a
ARR and 95% CI are estimated with no ANS (not exposed to ANS) group used as referent categoryemodels adjust for birthweight, sex, multiple births, small for gestational age, maternal variables
(age, marital status, race, diabetes, rupture of membranes
24 h, antepartum hemorrhage, and mode of delivery), center, and epoch; b Significant with P value <.05; c Model does not adjust for
centeremodel did not converge with center included.
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018.

SUPPLEMENTARY TABLE 3
Robust Poisson regression analysis without and with multiple imputation for selected outcomes with missing data
>1% or with imbalance of missing data >0.3% between groups
Without imputation With imputation
Outcome ARR (95% CI) ARR (95% CI)
Bronchopulmonary dysplasia [physiologic definition] or 0.94 (0.91e0.98) 0.95 (0.91e0.98)
death by 36 wk’ postmenstrual age
Population of survivors
Bronchopulmonary dysplasia, physiologic definition 0.96 (0.89e1.03) 0.97 (0.90e1.03)
Treatment with postnatal steroids for 0.98 (0.82e1.18) 0.96 (0.79e1.13)
bronchopulmonary dysplasia
Retinopathy of prematurity stage
3 or treated with 0.76 (0.66e0.87) 0.77 (0.66e0.87)
ablation/anti-VEGF druga
Respiratory support at discharge, oxygena 1.17 (1.03e1.33) 1.20 (1.05e1.35)
Prolonged hospital stay
120 d, all causes 0.94 (0.84e1.05) 0.95 (0.85e1.05)
Population of survivors eligible for follow-up
Respiratory support at 18e22 mo corrected age, 0.91 (0.42e1.99) 1.03 (0.21e1.86)
ventilation/CPAPa
Oxygen at 18e22 mo corrected agea 1.01 (0.65e1.58) 1.07 (0.60e1.54)
ARR and 95% CI are estimated with no antenatal corticosteroids (not exposed to antenatal corticosteroids) group used as referent categoryemodels adjust for birthweight, sex, multiple births, small
for gestational age, maternal variables (age, marital status, race, diabetes, rupture of membranes
24 h, antepartum hemorrhage, and mode of delivery), center, and epoch.
ARR, adjusted relative risk; CI, confidence interval; CPAP, continuous positive airway pressure; VEGF, vascular endothelial growth factor.
a
Model does not adjust for centeremodel did not converge with center included.
Travers et al. Antenatal corticosteroid exposure and extremely preterm infant outcomes. Am J Obstet Gynecol 2018.

130.e13 American Journal of Obstetrics & Gynecology JANUARY 2018