Purpose: Over the last 10 years, recommendations regarding the ideal level of oxygen for
resuscitation in preterm infants have changed from 100 percent, down to low levels of
oxygen (<30 percent), up to moderate concentration (30-65 percent). In addition, in 2010,
oxygen saturation targeting was recommended as standard of care and this contributed to a
change in clinical practice as clinicians were more likely and comfortable to start
resuscitation at either 21percent (room air) or titrated levels of oxygen such as 30-40
percent. When the guidelines were again revised in 2015, the International Liaison
Committee on Resuscitation (ILCOR) acknowledged that a critical knowledge gap continued
to exist for the resuscitation of the preterm infants <37 weeks, highlighting the need to
provide more concrete guidelines. This leaves clinicians in a challenging position.
Despite the advances that have been achieved in perinatal and neonatal care, neonates are
still vulnerable to the consequences of the oxidative effects from hyperoxia as well as
the deleterious effects from hypoxia. A large, multi-centre international trial of
sufficient sample size that is powered to look at safety outcomes such as mortality and
adverse neurodevelopmental outcomes is required to provide the necessary evidenced to
guide clinical practice with confidence.
Hypothesis: the null hypothesis for this study is that the incidence of mortality or
abnormal neurodevelopmental outcomes at 24+/- 6 months corrected age will be no different
by using either higher initial oxygen concentration of 60 percent compared to using lower
initial oxygen concentration of 30 percent for resuscitation of preterm infants of 23
0/7- 28 6/7 weeks gestation.
Justification:
The use of supplementary oxygen may be crucial, but also potentially detrimental to
premature infants at birth. High oxygen levels may lead to organ damage through oxidative
stress, while low oxygen levels may lead to increased mortality. Excess oxygen exposure
during the early post-birth period is associated with many complications and morbidities
of preterm birth. Preterm infants have lower levels of anti-oxidant pathways consistent
with their expected fetal environment of low oxygen exposure. Excess of oxygen
free-radicals in infants intrinsically deficient in enzymatic antioxidants and
non-enzymatic antioxidants may contribute to these morbidities. Pulmonary oxygen
toxicity, through the generation of reactive oxygen and nitrogen species in excess of
antioxidant defenses, is believed to be a major contributor to the development of
bronchopulmonary dysplasia (BPD). Using lower oxygen concentrations at birth results in
decreased oxidative stress markers and a decrease risk of developing BPD compared to
higher oxygen concentrations. Other organs that may be damaged by such oxidative stress
include kidneys, myocardium and the retina.
There is equally growing evidence that using lower oxygen concentrations will lead to
lower oxygen saturation levels and bradycardia, which may lead to increased rates of
mortality in this vulnerable group of infants. An individual patient analysis of clinical
trials reported that 46% of preterm infants resuscitated with initial low oxygen
concentration did not reach SpO2 of 80% at 5 min. This was associated with increased risk
of major intraventricular hemorrhage (IVH), and an almost five times higher risk of death
in this vulnerable group of infants. These data provide a warning note for the use of
higher vs. lower initial oxygen concentration during delivery room resuscitation. As the
investigator proceed in determining a safe range for resuscitation of ELBW/ELGA infants,
it is highly likely that the optimum level of oxygen concentration is between the two
extremes of 21 percent and 100 percent.
Objectives: To determine whether initial resuscitation of preterm neonates with 60
percent versus 30 percent oxygen results in better neurodevelopmental outcomes at 24+/- 6
months.
Research Method/Procedures: This will be a cluster crossover design, unmasked randomized
controlled trial (RCT) comparing two oxygen concentrations at initiation of
resuscitation. Infants will be placed on the resuscitation table with the initial steps
of resuscitation carried out as per standard of care at each centre which usually follows
current resuscitation guidelines. All centres will make every effort to establish
adequate lung expansion using CPAP or positive pressure ventilation as needed. Enrolled
infants will have a pulse oximeter sensor placed on the right arm in the first minute of
life. Their resuscitation will be initiated with an oxygen concentration of 30 or 60
percent depending on the randomization sequence at the centre at the given time. Infants
in the 30 percent group will remain in 30 percent oxygen until 5 min of age unless the
infant's heart rate (HR) remains 100/min or less and does not show a tendency towards
progressive increase before reaching 5 min of age or infant needs chest compression
and/or epinephrine. No alteration in oxygen concentration will be made for an infant who
is responding to resuscitation efforts with HR progressively increasing as minutes go by.
At 5 min of age, the clinical team will assess oxygen saturation. If the saturation is
less than 85 percent, oxygen should be increased by 10-20 percent every 60 sec to achieve
saturations of 85 percent or greater or a saturation of 90-95 percent at 10 min of age.
If saturations are greater than 95 percent at or before 5 min of age, oxygen should be
decreased stepwise (every 60 sec) with an aim to maintain saturations of 85 percent or
greater during 5-10 min of age or 90-95 percent at and beyond 10 min of age. The
procedure for infants in the 60 percent group will be identical. The intervention
duration for the trial will be the first 5 min after birth followed by initial
monitoring/action for the next 5 min where titration in oxygen concentration will be made
to achieve stability making a total of 10 min for study intervention. Titration of oxygen
before 5 min after birth will only be made if the infant remains bradycardic (HR less
than 100) and does not show a tendency towards a sustained increase in HR or if the
oxygen saturation exceeds 95 percent. If the infant does not respond to ventilation with
increasing HR in the first 5 min after birth, steps to ensure effective ventilation
should be done before oxygen is titrated.
Plan for Data Analysis: Generalized linear mixed model with binary outcome and maximum
likelihood estimate will be used to evaluate the effect of an oxygen concentration on the
primary outcome (as a composite at 24+/- 6 months corrected age of all-cause mortality or
the presence of a major neurodevelopmental outcome). To account for cluster crossover
design of the study, effects of centers (clusters) and a period (oxygen concentration)
within center will be considered random, and effects of a period (oxygen concentration)
will be entered as a fixed effect. This hierarchical model allows for the correlation of
patients within periods and within clusters. The model will be adjusted for gestational
age and whether or not infant required mask ventilation as potential confounding
variables. Similar generalized linear mixed models will be performed to evaluate the
effect of group on secondary outcomes. In addition, three subgroup analysis will be
performed: i) Gestational age will be categorized into 2 categories: 23+0- 25+6 vs.
26+0-28+6 weeks; ii) Breathing support will be categorized by infants supported only with
CPAP vs. received mask ventilation; iii) Sex/Gender will be categorized into 2
categories: female vs. male. For subgroup analysis baseline characteristics will be
compared using linear and generalized linear mixed models. Sensitivity analysis will be
performed to analyze the missing data; however, a very low number of missing values are
expected due to the design of the study.
