ARTICLE

Resuscitation of Preterm Neonates by Using Room

Air or 100% Oxygen
Casey L. Wang, MDa, Christina Anderson, MDb, Tina A. Leone, MDa, Wade Rich, RRTa, Balaji Govindaswami, MBBS, MPHb, Neil N. Finer, MDa

aDepartment of Pediatrics, Division of Neonatology, University of California, San Diego, California; bSanta Clara Valley Medical Center, San Jose, California

The authors have indicated they have no financial relationships relevant to this article to disclose.

What’s Known on This Subject What This Study Adds

Current guidelines for initial FIO2 during resuscitation of very preterm infants include the The current study is, to our knowledge, the first prospective, randomized comparison of
full range of available FIO2 values, whereas the best available evidence indicates a po- the use of room air versus oxygen for the initial resuscitation of very preterm infants and
tential advantage to the use of room air. raises concerns regarding the safety of room air for this population.

ABSTRACT
OBJECTIVE. In this study of preterm neonates of ⬍32 weeks, we prospectively compared
the use of room air versus 100% oxygen as the initial resuscitation gas.
www.pediatrics.org/cgi/doi/10.1542/
METHODS. A 2-center, prospective, randomized, controlled trial of neonates with peds.2007-1460
gestational ages of 23 to 32 weeks who required resuscitation was performed. The doi:10.1542/peds.2007-1460
oxygen group was initially resuscitated with 100% oxygen, with decreases in the This trial has been registered at
fraction of inspired oxygen after 5 minutes of life if pulse oxygen saturation was www.clinicaltrials.gov (identifier
⬎95%. The room air group was initially resuscitated with 21% oxygen, which NCT00369720).
was increased to 100% oxygen if compressions were performed or if the heart Key Words
rate was ⬍100 beats per minute at 2 minutes of life. Oxygen was increased in oxygen, pulse oximeter, resuscitation,
room air, very low birth weight
25% increments if pulse oxygen saturation was ⬍70% at 3 minutes of life or
Abbreviations
⬍80% at 5 minutes of life. SpO2—pulse oxygen saturation
NRP—Neonatal Resuscitation Program
RESULTS. Twenty-three infants in the oxygen group (mean gestational age: 27.6 weeks; PPV—positive pressure ventilation
range: 24 –31 weeks; mean birth weight: 1013 g; range: 495–2309 g) and 18 in the FIO2—fraction of inspired oxygen
room air group (mean gestational age: 28 weeks; range: 25–31 weeks; mean birth UCSD—University of California, San Diego
weight: 1091 g; range: 555–1840 g) were evaluated. Every resuscitated patient in the Accepted for publication Sep 25, 2007
room air group met rescue criteria and received an increase in the fraction of inspired Address correspondence to Neil Finer, MD,
402 Dickinson St, MPF Building, Suite 1-140,
oxygen by 3 minutes of life, 6 patients directly to 100% and 12 with incremental San Diego, CA 92103-8774. E-mail: nfiner@
increases. Pulse oxygen saturation was significantly lower in the room air group from ucsd.edu
2 to 10 minutes (pulse oxygen saturation at 3 minutes: 55% in the room air group PEDIATRICS (ISSN Numbers: Print, 0031-4005;
vs 87% in the oxygen group). Heart rates did not differ between groups in the first Online, 1098-4275). Copyright © 2008 by the
American Academy of Pediatrics
10 minutes of life, and there were no differences in secondary outcomes.
CONCLUSIONS. Resuscitation with room air failed to achieve our target oxygen saturation by 3 minutes of life, and we
recommend that it not be used for preterm neonates. Pediatrics 2008;121:1083–1089

R OOM AIR RESUSCITATION has been used successfully for asphyxiated term neonates. Multiple trials showed room
air to be at least as effective as 100% oxygen for resuscitation in this population.1–6 Meta-analyses of the major
randomized, controlled trials completed to date showed improved survival rates for term infants resuscitated with
room air, compared with 100% oxygen. Subgroup analysis showed a larger survival benefit for infants born at ⬍37
weeks of gestation.7–9 Despite this information suggesting that room air might be beneficial for resuscitating preterm
infants, there has never been a trial specifically evaluating the use of room air versus higher oxygen concentrations
during newborn resuscitation in this group of infants.
The World Health Organization has stated that supplemental oxygen is not necessary as the initial resuscitating
gas.10,11 The International Liaison Committee on Resuscitation systematically reviewed the available evidence for the
use of oxygen during newborn resuscitation and concluded that insufficient evidence existed to specify the
concentration of oxygen that should be used at the onset of resuscitation.12 The Neonatal Resuscitation Program
(NRP) published the most recent revision of its textbook in 2006, which continues to recommend using 100%
oxygen as the initial ventilating gas for term infants.13 In addressing the care of preterm infants, this edition of the
NRP textbook does not recommend a specific initial concentration of oxygen but does recommend monitoring pulse
oxygen saturation (SpO2) levels with pulse oximetry in the delivery room.
The potential for toxicity resulting from hyperoxia in preterm neonates has long been hypothesized. There is now
an emerging body of information suggesting that many of the morbid conditions associated with extreme immaturity

PEDIATRICS Volume 121, Number 6, June 2008 1083

Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015
are potentiated by an excess of free radicals occurring in admitting ⬃110 very low birth weight infants per year,
infants who are intrinsically deficient in antioxidants, and at Santa Clara Valley Medical Center, a 40-bed
such as superoxide dismutase, catalase, and glutathione regional NICU admitting ⬃80 very low birth weight
peroxidase.14 Low plasma antioxidant activity at birth in infants per year. This project was approved by the insti-
preterm infants has been shown to be an independent tutional review boards of both centers, and prenatal
risk factor for death.14 Associations have been made parental consent was obtained for all enrolled subjects.
0
between hyperoxia and important neonatal outcomes, Subjects were all inborn, with gestational ages of 23 ⁄7 to
6 ⁄7
including necrotizing enterocolitis, retinopathy of pre- 31 weeks. Neonates with known congenital malforma-
maturity, and bronchopulmonary dysplasia.15–20 A num- tions or chromosomal anomalies were excluded. Multi-
ber of animal models21–23 have demonstrated tissue dam- ple gestations were assigned randomly according to
age from reactive oxygen species triggered during pregnancy, not individual neonates. When delivery was
resuscitation with 100% oxygen. End-organ effects, in- imminent, subjects were randomly assigned to 2 groups
cluding elevated cardiac enzyme levels, have been doc- by using sealed envelopes. Randomization was in blocks
umented in term neonates resuscitated with 100% ox- of 10. Patients were included in the trial if any resusci-
ygen.24 None of these observations is the result of tation was received. The study was designed to test
prospective, randomized, clinical trials. feasibility, with the goals of evaluating SpO2 values dur-
It is well known that intrauterine PaO2 levels are 15 to ing resuscitation in the 2 groups, determining whether
25 mm Hg before delivery25 and increase to 50 to 80 there was significant separation between the groups, and
mm Hg after delivery. Data obtained from a randomized evaluating whether the interventions were safe.
trial comparing the use of room air or 100% oxygen for At UCSD Medical Center, each delivery was attended
the resuscitation of asphyxiated term infants reported by a team consisting of a pediatric resident, a neonatal
arterial blood gas levels obtained from the umbilical fellow, a neonatal nurse, and a respiratory therapist. An
artery during resuscitation. In the infants resuscitated attending neonatologist also was present at most of the
with pure oxygen, the PaO2 levels were ⬎100 mm Hg at deliveries. Colorimetric carbon dioxide detectors (Pedi-
5 to 6 minutes, 15 minutes, and thereafter, whereas cap; Nellcor Puritan Bennett, Pleasanton, CA) were used
values for infants resuscitated with room air did not to facilitate bag-mask ventilation29 and to confirm intu-
exceed 80 to 90 mm Hg (126.3 ⫾ 21.8 mm Hg at 6.8 bation. A T-piece resuscitator (Neopuff; Fisher&Paykel,
minutes of life vs 72.2 ⫾ 6.8 mm Hg at 5.3 minutes of Auckland, NZ) was used in most cases, with occasional
life).4 use of a flow-inflating bag. At Santa Clara Valley Medical
Observations in healthy neonates in the first minutes Center, all deliveries were attended by a team consisting
after birth demonstrate a slow increase in SpO2 levels of a neonatal respiratory therapist and either a neonatal
over the first 15 minutes of life, with a median SpO2 at 3 nurse practitioner, a pediatric resident, or a pediatrician.
minutes of life of 76% (interquartile range: 64%– 87%) A neonatologist also attended deliveries of infants at
and a median SpO2 at 5 minutes of life of 90% (79%– ⬍28 weeks of gestation. Flow-inflating bags were used.
91%). In the few preterm neonates evaluated, the me- For both groups at both centers, routine NRP protocol
dian SpO2 at 5 minutes of life was 83% (74%–91%).26 was followed.
The median time to reach a SpO2 of ⬎90% has been For all infants, a pulse oximeter was applied by the
reported to be 8 minutes.27 A significant gradient be- respiratory therapist within the first 30 seconds of life.
tween preductal and postductal SpO2 persists for the first All SpO2 values obtained and recorded were from a
15 minutes of life, with most untreated neonates not preductal site, always the right wrist. The base unit was
reaching a preductal SpO2 of 90% in the first 5 minutes turned on and the probe was applied to the infant,
of life.28 Limited information exists on the SpO2 of pre- followed by attachment of the probe to the base unit.
term infants during resuscitation, and studies to date This method follows manufacturer’s guidelines, to de-
have not evaluated SpO2 with different delivered frac- crease maximally the time to a reliable signal, as con-
tions of inspired oxygen (FIO2) values in this population. firmed by O’Donnell et al.30 Radical oximeters (Masimo,
We initiated this trial because, despite the suggestions Irvine, CA) were used with HiFi sensors (Masimo, Ir-
that room air would be a better resuscitation gas for vine, CA), which automatically set the oximeter to max-
preterm infants, the use of room air for resuscitation has imal sensitivity, and 2-second averaging. Resuscitations
never been evaluated systematically in this special pop- at UCSD were video-recorded by using our previously
ulation, especially the smallest preterm infants. We hy- described methods.31 A purpose-built, computerized,
pothesized that, for preterm neonates, resuscitation ini- data acquisition system continuously collected delivered
tiated with room air and adjusted on the basis of SpO2 FIO2 data by using the output of an inline polarographic
levels (targeted oxygen delivery) would be more effec- oxygen analyzer, airway pressure from the T-piece or
tive than pure oxygen in achieving SpO2 values similar to flow-inflating bag, time-linked video, and pulse oxime-
those of nonresuscitated transitioning neonates. ter outputs, including pulse rate and SpO2.32 At Santa
Clara Valley Medical Center, the method was similar,
METHODS with the following exceptions: video recordings and
This was a prospective, dual-center, randomized, con- computerized data acquisition were not used. Data were
trolled, clinical trial conducted at the University of Cal- obtained from 2 independent observers and pulse
ifornia, San Diego (UCSD), Medical Center, a regional oximeter downloads.
NICU with 40 beds and a high-risk perinatal service For the oxygen group, resuscitation was initiated with

1084 WANG et al
Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015
 TABLE 1 Maternal Factors and Baseline Characteristics
Room Air 100% Oxygen
(n ⫽ 18) (n ⫽ 23)
Gestation, mean ⫾ SD, wk 28.1 ⫾ 2.23 27.6 ⫾ 2.1
Birth weight, mean ⫾ SD, g 1066 ⫾ 368 1013 ⫾ 444
Female, n (%) 11 (61) 14 (61)
Prenatal steroid therapy, n (%) 11 (62) 17 (74)
IUGR, n (%) 3 (17) 8 (35)
Maternal age, mean ⫾ SD, y 28 ⫾ 8.5 28 ⫾ 6.5
PPROM, n (%) 5 (28) 10 (43)
Cesarean section, n (%) 9 (50) 16 (70)
Maternal chorioamnionitis, n (%) 3 (17) 8 (35)
Singleton, n (%) 16 (89) 16 (69)
PIH, n (%) 6 (33) 5 (21)
IUGR indicates intrauterine growth retardation; PPROM, preterm prolonged rupture of mem-
branes; PIH, pregnancy-induced hypertension.
FIGURE 1
Protocol for room air resuscitation. HR indicates heart rate; values are in beats per minute.
first dose of prenatal steroid treatment. The two patients
who were randomly assigned (1 to each group) but did
100% oxygen. At 5 minutes of life, oxygen treatment not require any resuscitation, were a 31-week preterm
was weaned if the SpO2 was consistently ⬎95%. In the neonate assigned to the oxygen group and a 29-week
room air group, 21% oxygen was used as the initial preterm infant assigned to the room air group. Neither
resuscitation gas. FIO2 was immediately increased to required continuous positive airway pressure therapy or
100% under the following conditions: need for chest PPV and these patients were not included in the analy-
compressions or medication administration, heart rate of ses.
⬍100 beats per minute at 2 minutes of life, or heart rate
of ⬍60 beats per minute for 30 seconds at any time. FIO2 Resuscitations
was increased in 25% increments if SpO2 was ⬍70% at Every patient in the room air group required an increase
3 minutes of life or ⬍85% at 5 minutes of life. This in FIO2 at or before 3 minutes of life. FIO2 was increased
method is detailed in Fig 1. directly to 100% because of bradycardia by 2 minutes of
The research team reviewed the available video re- age for 6 patients, and FIO2 was increased incrementally
cordings and evaluated adherence to the NRP and study for failure to meet SpO2 criteria at 3 minutes of life for
protocols. With the video recordings linked to the analog the remaining 12 patients. Heart rates of ⬍100 beats per
data on heart rate, SpO2, FIO2, peak inspiratory pressure, minute at 2 minutes of life were seen in 4 patients in the
and positive end expiratory pressure, the exact time at oxygen group. Significant differences were seen in SpO2
which resuscitation events occurred, including any at 2, 3, 4, 5, 6, 7, 8, 9, and 10 minutes (P ⫽ .01, analysis
changes in FIO2 and administration of positive pressure of variance) (Fig 2). The delivered FIO2 is detailed in Fig
ventilation (PPV), was documented in relation to the 3. Heart rates did not differ between groups in the first
status of the infant at the time.
Descriptive statistics were calculated by using SigmaStat
3.0.1a (Systat, San Jose, CA). Student’s t test was used to
compare normally distributed variables. The Mann-Whit- TABLE 2 Delivery Room Interventions and Early Parameters
ney rank-sum test was used to test significance for non– Room Air 100% Oxygen P
normally distributed variables. One-way, repeated-mea- (n ⫽ 18) (n ⫽ 23)
sures analysis of variance was performed for SpO2 by using Surfactant treatment in DR, n (%) 10 (55) 10 (43) .5
SPSS for Windows 10 (SPSS, Chicago, IL). Intubation in DR, n (%) 10 (55) 10 (43) .5
Chest compressions in DR, n (%) 0 (0) 3 (13) .48
RESULTS Medications in DR, n (%) 0 (0) 1 (4) .8
Apgar score at 1 min, median 5 4 .8
Baseline Characteristics Apgar score at 5 min, median 8 9 .034
Forty-three infants were randomized in this trial be- Apgar score at 10 minutes, mediana 8 7 .43
tween December 2005 and March 2007, two of whom Cord arterial pH, mean ⫾ SD 7.3 ⫾ 0.05 7.27 ⫾ 0.08 .2
did not require resuscitation. Eighteen infants received PPV, n (%) 16 (89) 22 (95) .7
room air and 23 infants received oxygen. A total of 32 CPAP only, n (%) 2 (11) 1 (5) .96
Initial arterial blood gas levels after
patients were enrolled at UCSD and 11 at Santa Clara
resuscitation
Valley Medical Center. Baseline characteristics and ma- pH, mean ⫾ SD 7.26 ⫾ 0.18 7.3 ⫾ 0.09 .44
ternal factors are detailed in Table 1. No differences were PCO2, mean ⫾ SD, mm Hg 54 ⫾ 31 46 ⫾ 7 .9
seen between groups with respect to baseline character- PO2, mean ⫾ SD, mm Hg 59 ⫾ 16 68 ⫾ 28 .4
istics, delivery room interventions (Table 2), or maternal Base deficit, mean ⫾ SD, mol 1.5 ⫾ 5.7 4 ⫾ 3.1 .17
factors. Prenatal steroid administration was defined as DR indicates delivery room; CPAP, continuous positive airway pressure.
delivery of the neonate a minimum of 48 hours after the a Apgar scores at 10 minutes were determined for only 5 patients in each group.

PEDIATRICS Volume 121, Number 6, June 2008 1085

Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015
FIGURE 2 FIGURE 4
Mean SpO2 at each minute of life. Bars represent SD. Mean heart rate at each minute of life. Values are in beats per minute.

10 minutes of life, and the times to establish a heart rate DISCUSSION

of ⬎100 beats per minute did not differ (Fig 4). A greater To our knowledge, this is the first prospective random-
percentage of neonates in the oxygen group had SpO2 of ized study comparing the use of room air and pure
⬎95% at each minute of resuscitation (P ⬍ .05 at 4 and oxygen in preterm infants of gestational age of ⬍32
5 minutes only, Mann-Whitney test) (Fig 5). weeks. In designing our trial, we attempted to compare
the standard of care for resuscitation with a new targeted
Protocol and NRP Deviations oxygen delivery approach. The most current edition of
Two patients in the room air group had their FIO2 in- the NRP textbook indicates that resuscitation of preterm
creased because of bradycardia before initiation of PPV. neonates can be initiated with FIO2 between 21% and
This represents a deviation from both NRP guidelines 100% and that SpO2 should be monitored. Our study
and our protocol. One patient in the room air group with was initiated before the introduction of this recommen-
SpO2 in the 30% to 50% range did not have FIO2 in- dation, at a time when most centers in the United States
creased until 7 minutes of life. Two patients in the oxy- resuscitated preterm neonates with 100% oxygen and
gen group had their oxygen weaned at 4 minutes of life. did not have the capability to provide blended oxygen.
In addition, most centers did not routinely use pulse
Secondary Outcomes oximeters in the delivery room.33 Therefore, most pre-
Secondary outcomes are detailed in Table 3. There were no term infants resuscitated in the United States would
significant differences in the occurrence of any of the eval- have received 100% oxygen for the entire duration of
uated outcomes, including intraventricular hemorrhage, the delivery room stay. We have shown that this dura-
retinopathy of prematurity, necrotizing enterocolitis, and tion in our center is 23 minutes, on average.34 We chose
chronic lung disease. There was 1 death in each group. In to use 100% oxygen in the control group for the first 5
the room air group, an infant of 24 weeks of gestation died minutes of life because this was the standard of care at
at 3 days of life as a result of respiratory failure, pulmonary the time. We chose to initiate resuscitation with room air
hemorrhage, and grade IV intraventricular hemorrhage; in in the study group because of the evidence from studies
the oxygen group, an infant of 25 weeks of gestation died in term neonates that room air is as effective as or better
at 7 days of life as a result of sepsis, respiratory distress
syndrome, and pneumothorax.

FIGURE 3 FIGURE 5
Mean level of oxygen administered at each minute of life. Bars represent SD. Proportion of patients with SpO2 of ⬎95% at each minute of life.

1086 WANG et al
Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015
 TABLE 3 Secondary Outcomes the infants failed to meet the target with room air as the
Room Air 100% Oxygen P resuscitating gas. The actual SpO2 levels that are too high
(n ⫽ 18) (n ⫽ 23) or too low for preterm neonates, and therefore unsafe,
are not known. After the transitional period, most NICUs
Death, n (%) 1 (6) 1 (4) .95
Grade III–IV IVH, n (%) 2 (11) 0 .55 set upper and lower SpO2 limits for infants receiving
Oxygen therapy at adjusted age of 36 wk, n (%) 7 (39) 3 (13) .13 oxygen therapy. Although neonatologists would likely
Isolated gastrointestinal perforation, n (%) 1 (8) 2 (9) .9 agree on the need for setting such limits, agreement
Duration of intubation, median ⫾ SD, d 2 ⫾ 16.8 1 ⫾ 13.3 .51 among units on any actual number is doubtful.
Duration of NCPAP therapy, mean ⫾ SD, d 11 ⫾ 13 12 ⫾ 16 .47 One could question whether allowing a longer inter-
Pneumothorax, n (%) 0 3 (13) .48 val before such rescue would have resulted in sponta-
IVH indicates intraventricular hemorrhage; NCPAP, nasal continuous positive airway pressure. neously improving SpO2 values. We think that this is
unlikely; 1 infant in the room air group was not given
oxygen until 7 minutes of life because the team was not
than pure oxygen. We hypothesized that the best FIO2 certain that the oximeter was functioning. The infant did
for resuscitation might not be the extremes of either not have bradycardia and seemed to be in otherwise
21% or 100% and that a targeted approach might be stable condition. This infant’s SpO2 values remained be-
most successful. We chose our SpO2 targets on the basis low 50% until the FIO2 was increased, at which point the
of the best available pulse oximeter data from observa- SpO2 rapidly increased.
tions of term and near term infants who did not require The need for a blender in the delivery area for the
active resuscitation at birth. We thought that the trajec- resuscitation of preterm infants highlights the poten-
tory of increasing SpO2 after birth in nonresuscitated tial for error, in that the team needs to check the
newborn infants would be the best available model to actual blender setting before the beginning of any
mimic during resuscitation. resuscitation, adding another variable to resuscitation
The SpO2 of a fetus in relatively stable condition is preparation and management. We have experienced
⬃50%35,36 but may be less at the time of birth. The the situation in which a nonstudy infant who was
transition to higher SpO2 after birth has been observed thought to be receiving 100% was actually receiving
by several investigators, mostly evaluating term nonre- room air. Another possible hazard in resuscitation
suscitated neonates. House et al37 studied 100 newborn with room air is the perceived need to compensate for
infants (weight: 850 –5230 g) delivered vaginally or low SpO2 by increasing ventilation, leading to possible
through cesarean section. The average arterial oxygen volutrauma or barotrauma.39
saturation was 59% at 1 minute, 68% at 2 minutes, 82% We postulate that the persistently low initial SpO2
at 5 minutes, and 90% at 15 minutes. Toth et al38 studied values we observed during room air resuscitation of very
50 healthy, vaginally delivered, newborn infants and preterm neonates may be related to the lack of adapta-
compared the SpO2 values from preductal and postductal tion of the pulmonary vasculature at birth in the absence
sites. Two minutes after birth, the mean preductal SpO2 of supplemental oxygen. Therefore, we think that the
was 73% (range: 44%–95%) and the mean postductal low SpO2 values probably reflect a fetal circulation with
SpO2 was 67% (range: 34%–93%). SpO2 levels of ⬎95% right-to-left shunting at the ductal and foraminal levels,
were reached after 12 minutes (range: 2–55 minutes) for secondary to continuing pulmonary vasoconstriction.
preductal values and after 14 minutes (range: 3–55 min- The gradient between preductal and postductal SpO2
utes) for postductal values.38 More recently, Kamlin et values in preterm neonates may be larger and persist
al26 reported SpO2 values in the first minutes after birth longer than reported by Mariani et al28 for the term
in healthy nonresuscitated neonates. The median level population. We speculate that there are different sensi-
at 3 minutes was 76% (interquartile range: 64%– 87%). tivities of the pulmonary circulation to oxygen in pre-
At 5 minutes of life, the median level was 80% (inter- term infants and term infants, and we think that this
quartile range: 40%–95%). This study demonstrated mechanism requires additional study in relevant animal
that infants of ⬍37 weeks who did not receive resusci- models. In addition, although we did not observe any
tation required 4.4 minutes to achieve SpO2 of 75% and significant difference in intrauterine growth retardation
7.3 minutes to achieve SpO2 of 90%. between the groups, there were fewer such cases in the
The targets we chose for adjustment of delivered FIO2 room air group, and fetal distress and/or hypoxia in
in the study group were near the median levels for utero, as may occur in infants with intrauterine growth
nonresuscitated neonates. These were the lowest levels retardation, may actually encourage this physiologic re-
we felt comfortable allowing with this new approach to sponse and require oxygen to reduce the elevated pul-
adjusting delivered FIO2. We were unable to meet these monary vascular pressures at birth.
targets in any resuscitated infants with room air used as The rate of PPV in our population was high, although
the initial gas. It may be argued that these targets were cord blood gas values on average did not suggest that the
too high and we failed to accomplish resuscitation with infants were compromised before delivery. We think
room air because we set the wrong targets. If our study that this finding reflects the increased difficulties that
population were similar to the populations used in the preterm infants have with the transition to neonatal life,
observational studies, with the median as the target, compared with term neonates. Our rate of PPV might be
then by definition 50% of the infants would not achieve higher than previously thought because of the objective
the target. In our study population, however, 100% of nature of our video review of the resuscitations and

PEDIATRICS Volume 121, Number 6, June 2008 1087

Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015
because we did not rely on recall to document resusci- REFERENCES
tation events. 1. Ramji S, Ahuja S, Thirupuram S, Rootwelt T, Rooth G, Saug-
Preliminary data from a recent prospective trial by stad OD. Resuscitation of asphyxic newborn infants with room
Escrig et al40 in which extremely low gestational age neo- air or 100% oxygen. Pediatr Res. 1993;34(6):809 – 812
2. Saugstad OD, Rootwelt T, Aalen O. Resuscitation of asphyxi-
nates were randomly assigned to resuscitation with 100%
ated newborn infants with room air or oxygen: an interna-
or ⬍40% oxygen showed that both groups achieved the
tional controlled trial: the Resair 2 study. Pediatrics. 1998;
target SpO2 of 85% at ⬃8 minutes of age. In view of these 102(1). Available at: www.pediatrics.org/cgi/content/full/102/
observations and our own results, we think that a starting 1/e1
FIO2 of 30% to 40% is appropriate for resuscitation of 3. Vento M, Asensi M, Sastre J, Garcı́a-Sala F, Pallardó FV, Viña J.
preterm neonates. Subsequent adjustment of FIO2 may be Resuscitation with room air instead of 100% oxygen prevents
necessary to allow the SpO2 values to increase slowly to oxidative stress in moderately asphyxiated term neonates. Pe-
85% by 7 to 10 minutes of age. diatrics. 2001;107(4):642– 647
This trial is somewhat limited by the small sample 4. Vento M, Sastre J, Asensi M, Lloret A, Garcı́a-Sala F, Viña J.
size. However, the uniformity of results from 2 indepen- Oxidative stress in asphyxiated term infants resuscitated with
100% oxygen. J Pediatr. 2003;142(3):240 –246
dent centers suggests that the findings were reliable and
5. Ramji S, Rasaily R, Mishra PK, et al. Resuscitation of asphyx-
reproducible. Every resuscitated infant assigned ran-
iated newborns with room air or 100% oxygen at birth: a
domly to room air at either site received an increase in multicentric clinical trial. Indian Pediatr. 2003;40(6):510 –517
FIO2 to achieve the targeted SpO2. Our study population 6. Bajaj N, Udani RH, Nanavati RN. Room air versus 100 percent
included a wide range of gestational ages, from 23 to 32 oxygen for neonatal resuscitation: a controlled clinical trial. J
weeks. Although there may be differences at the ex- Trop Pediatr. 2005;51(4):206 –211
tremes of gestational ages, our sample size was not large 7. Saugstad OD, Ramji S, Vento M. Resuscitation of depressed
enough for reliable evaluation of subgroups. The study newborn infants with ambient air or pure oxygen: a meta-
population of infants of ⬍32 weeks of gestation was analysis. Biol Neonate. 2005;87(1):27–34
consistent with the population of preterm infants for 8. Davis PG, Tan A, O’Donnell CPF, Schulze A. Resuscitation of
newborn infants with 100% oxygen or air: a systematic review
whom the NRP created new recommendations.
and meta-analysis. Lancet. 2004;364(9442):1329 –1333
The trial was not blinded, by design. In part, this was
9. Rabi Y, Rabi D, Yee W. Room air resuscitation of the depressed
because of the need for additional personnel in the de- newborn: a systematic review and meta-analysis. Resuscitation.
livery room. The lack of blinding might have resulted in 2007;72(3):353–363
differential resuscitation, including the possibility that 10. World Health Organization. Basic Newborn Resuscitation: A Prac-
the room air-resuscitated infants received more-aggres- tical Guide. Geneva, Switzerland: World Health Organization;
sive ventilation. However, 31 infants were enrolled at a 1998
site that uses continuous video recording linked with 11. World Health Organization. Basic Newborn Resuscitation: A Prac-
analog data on heart rate, FIO2, inspired pressure, and tical Guide, Chapter 2. Geneva, Switzerland: World Health
SpO2. This method of obtaining data allows for accurate Organization; 1998. Available at: www.who.int/reproductive-
health/publications/newborn_resuscitation/chap2.html. Ac-
unbiased timing of events and analysis of interventions
cessed September 24, 2007
and physiologic responses. Although the intervention
12. Niermeyer S, Kattwinkel J, VanReempts P, et al. International
was not blinded, the analysis of data was performed in guidelines for neonatal resuscitation: an excerpt from the
the most objective way possible. Because of the small Guidelines 2000 for Cardiopulmonary Resuscitation and Emer-
size of the trial, we did not perform a prospective power gency Cardiovascular Care: International Consensus on Sci-
calculation. On the basis of the SpO2 differences at 3 ence. Pediatrics. 2000;106(3). Available at: www.pediatrics.org/
minutes (55.71 ⫾ 17.3% for the room air group and cgi/content/full/106/3/e29
87.39 ⫾ 7.57% for the oxygen group), the actual power 13. Kattwinkel J, ed. Textbook of Neonatal Resuscitation. 5th ed. Elk
of the trial was 95%. However, this study did not have Grove Village, IL: American Academy of Pediatrics; 2006
adequate power to assess the impact of the intervention 14. Silvers KM, Gibson AT, Russell JM, Powers HJ. Antioxidant
activity, packed cell transfusions, and outcome in premature
on long-term outcomes.
infants. Arch Dis Child Fetal Neonatal Ed. 1998;78(3):F214 –F219
In view of our observations, we recommend that,
15. Tin W, Milligan DW, Pennefather P, Hey E. Pulse oximetry,
except for current or future clinical trials, room air severe retinopathy, and outcome at one year in babies of less
should not be used as an initial resuscitating gas for than 28 weeks gestation. Arch Dis Child Fetal Neonatal Ed. 2001;
preterm neonates of ⬍32 weeks. Additional studies are 84(2):F106 –F110
required to determine the ideal initial FIO2 for resuscita- 16. Saugstad OD. Bronchopulmonary dysplasia and oxidative
tion of these infants. stress: are we closer to an understanding of the pathogenesis of
BPD? Acta Paediatr. 1997;86(12):1277–1282
17. Luo XP, Jankov RP, Ning Q, Liao LJ, Tanswell AK. Oxygen-
ACKNOWLEDGMENTS mediated parenchymal and vascular lung injury. Acta Pharma-
col Sin. 2002;23(suppl):22–28
We thank the neonatal staff members at both participat-
18. Davis JM. Role of oxidant injury in the pathogenesis of neo-
ing institutions. Without the support of our neonatal
natal lung disease. Acta Paediatr Suppl. 2002;91(437):23–25
nurses, neonatal respiratory therapists, and pediatric res- 19. Bell EF. Preventing necrotizing enterocolitis: what works and
idents, this trial would not have been possible. We also how safe? Pediatrics. 2005;115(1):173–174
thank Glenn DeSandre, MD, Anita Sit, MD, and Claire 20. Bisquera JA, Cooper TR, Berseth CL. Impact of necrotizing
Carroll, RN, MSN. We thank the families of these neo- enterocolitis on length of stay and hospital charges in very low
nates for allowing their infants’ participation in this trial. birth weight infants. Pediatrics. 2002;109(3):423– 428

1088 WANG et al
Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015
21. Poulsen JP, Oyasaeter S, Saugstad OD. Hypoxanthine, xan- 32. Roberts KD, Leone TA, Edwards WH, Rich WD, Finer NN.
thine, uric acid in newborn pigs during hypoxemia followed by Premedication for nonemergent neonatal intubations: a ran-
resuscitation with room air or 100% oxygen. Crit Care Med. domized, controlled trial comparing atropine and fentanyl to
1993;21(7):1058 –1065 atropine, fentanyl, and mivacurium. Pediatrics. 2006;118(4):
22. Rootwelt T, Loberg EM, Moen A, Oyasaeter S, Saugstad OD. 1583–1591
Hypoxemia and reoxygenation with 21% or 100% oxygen in 33. Leone TA, Rich W, Finer NN. A survey of delivery room
newborn pigs: changes in blood pressure, base deficit, and resuscitation practices in the United States. Pediatrics. 2006;
hypoxanthine and brain morphology. Pediatr Res. 1992;32(1): 117(2). Available at: www.pediatrics.org/cgi/content/full/117/
107–113 2/e164
23. Rootwelt T, Odden J-P, Hall C, Saugstad OD. Cerebral blood 34. Wang CL, Leone TA, Rich W, Finer NN. Neonatal resuscitation
flow and evoked potentials during reoxygenation with 21% or of ELBW infants: time spent in the delivery room. E-PAS2006:
100% oxygen in newborn pigs. J Appl Physiol. 1993;75(5): 2860.205. Available at: www.abstracts2view.com/pasall/. Ac-
2054 –2060 cessed September 24, 2007
24. Vento M, Sastre J, Asensi MA, Vina J. Room-air resuscitation 35. Nijland R, Jongsma HW, Nijhuis JG, van den Berg PP, Oese-
causes less damage to heart and kidney than 100% oxygen. burg B. Arterial oxygen saturation in relation to metabolic
Am J Respir Crit Care Med. 2005;172(11):1393–1398
acidosis in fetal lambs. Am J Obstet Gynecol. 1995;172(3):
25. Kirschbaum TH, Lucas WE, DeHaven JC, Assali NS. The dy-
810 – 819
namics of placental oxygen transfer. Am J Obstet Gynecol. 1967;
36. Dildy GA, van den Berg PP, Katz M, et al. Intrapartum fetal
98(3):429 – 443
pulse oximetry: fetal oxygen saturation trends during labor and
26. Kamlin COF, O’Donnell CPF, Davis PG, Morley CJ. Oxygen
relation to delivery outcome. Am J Obstet Gynecol. 1994;171(3):
saturation in healthy infants immediately after birth. J Pediatr.
679 – 684
2006;148(5):585–589
37. House JT, Schultetus RR, Gravenstein N. Continuous neonatal
27. Rabi Y, Yee W, Chen SY, Singhal N. Oxygen saturation trends
immediately after birth. J Pediatr. 2006;148(5):590 –594 evaluation in the delivery room by pulse oximetry. J Clin Monit.
28. Mariani G, Dik PB, Ezquer A, et al. Pre-ductal and post-ductal 1987;3(2):96 –100
O2 saturation in healthy term neonates after birth. J Pediatr. 38. Toth B, Becker A, Seelbach-Gobel B. Oxygen saturation in
2007;150(4):418 – 421 healthy newborn infants immediately after birth measured by
29. Leone TA, Lange A, Rich W, Finer NN. Disposable colorimetric pulse oximetry. Arch Gynecol Obstet. 2002;266(2):105–107
carbon dioxide detector use as an indicator of a patent airway 39. Björklund LJ, Ingimarsson J, Curstedt T, Larsson A, Robertson
during noninvasive mask ventilation. Pediatrics. 2006;118(1). B, Werner O. Lung recruitment at birth does not improve lung
Available at: www.pediatrics.org/cgi/content/full/118/1/e202 function in immature lambs receiving surfactant. Acta Anaes-
30. O’Donnell CPF, Kamlin COF, Davis PG, Morley CJ. Obtaining thesiol Scand. 2001;45(8):986 –993
pulse oximetry data in neonates: a randomised crossover study 40. Escrig R, Arruza L, Izquierdo I, et al. Achievement of target
of sensor application techniques. Arch Dis Child Fetal Neonatal oxygen saturation in extremely low gestational neonates re-
Ed. 2005;90(1):F84 –F85 suscitated with different oxygen concentrations: a prospective
31. Carbine DN, Finer NN, Knodel E, Rich W. Video recording as a randomized clinical trial. E-PAS2007:617932.21. Available at:
means of evaluating neonatal resuscitation performance. Pedi- www.abstracts2view.com/pasall/. Accessed September 24,
atrics. 2000;106(4):654 – 658 2007

COURT UPHOLDS SCHOOL BAN ON CELL PHONES

“A ban on cell phones in the nation’s largest school system was upheld
Tuesday by a state appeals court. New York City’s Department of Education
passed rules in September 2005 barring students from having their phones in
public schools. School officials and Mayor Michael Bloomberg have called the
phones a distraction and say they could be used for nefarious purposes,
including cheating. Parents say they need to stay in touch with their children
in case of emergencies like the terrorist attacks of Sept. 11, 2001. They call the
ban irrational and unsafe and say it intrudes on their right to determine what
is best for their children. City lawyers argued that education officials had the
right to make policy decisions—‘the kind government officials make all the
time’—about devices students are allowed to have at school.”
Associated Press. Wall Street Journal. April 23, 2008
Noted by JFL, MD

PEDIATRICS Volume 121, Number 6, June 2008 1089

Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015
 Resuscitation of Preterm Neonates by Using Room Air or 100% Oxygen
Casey L. Wang, Christina Anderson, Tina A. Leone, Wade Rich, Balaji
Govindaswami and Neil N. Finer
Pediatrics 2008;121;1083
DOI: 10.1542/peds.2007-1460
Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/121/6/1083.full.h
tml
References This article cites 28 articles, 8 of which can be accessed free
at:
http://pediatrics.aappublications.org/content/121/6/1083.full.h
tml#ref-list-1
Citations This article has been cited by 23 HighWire-hosted articles:
http://pediatrics.aappublications.org/content/121/6/1083.full.h
tml#related-urls
Subspecialty Collections This article, along with others on similar topics, appears in
the following collection(s):
Fetus/Newborn Infant
http://pediatrics.aappublications.org/cgi/collection/fetus:newb
orn_infant_sub
Pulmonology
http://pediatrics.aappublications.org/cgi/collection/pulmonolo
gy_sub
Permissions & Licensing Information about reproducing this article in parts (figures,
tables) or in its entirety can be found online at:
http://pediatrics.aappublications.org/site/misc/Permissions.xht
ml
Reprints Information about ordering reprints can be found online:
http://pediatrics.aappublications.org/site/misc/reprints.xhtml

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

publication, it has been published continuously since 1948. PEDIATRICS is owned, published,
and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk
Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy of Pediatrics. All
rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015

 Resuscitation of Preterm Neonates by Using Room Air or 100% Oxygen
Casey L. Wang, Christina Anderson, Tina A. Leone, Wade Rich, Balaji
Govindaswami and Neil N. Finer
Pediatrics 2008;121;1083
DOI: 10.1542/peds.2007-1460

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/121/6/1083.full.html

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

publication, it has been published continuously since 1948. PEDIATRICS is owned,
published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point
Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy
of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org at Medical School Library on January 24, 2015