Resuscitation in the Delivery Room: Lung Protection From the First Breath

F_IO2 < 1.0

One of the goals of resuscitation is to provide sufficient oxygen perfusion of tissue. Oxygen is critical for life and is required for cell metabolism and growth. Even relatively short-term hypoxemia can have disastrous outcomes. Oxygen was first administered to a neonate in 1780. Since the 1940s, oxygen has been the most commonly used therapy in the delivery room and in the neonatal intensive care unit.⁸ High F_IO2 for a prolonged period (days to weeks) is a major factor in the pathogenesis of major newborn morbidity, including retinopathy of prematurity and bronchopulmonary dysplasia (BPD). High F_IO2 leads to the production of oxygen free radicals such as superoxide, hydrogen peroxide, and the hydroxyl radical. These are highly reactive chemical species that can react with and injure almost every type of molecule in living cells. High F_IO2 may prolong the time to spontaneous crying and breathing, increase oxygen consumption, decrease minute ventilation, cause atelectasis, or alter cerebral circulation (decrease cerebral blood flow).^9–17 By contrast, too low an F_IO2 can cause persistent pulmonary hypertension of the newborn, hypoxic brain injury, multisystem organ dysfunction, and is more likely to keep the ductus arteriosus open.

Several trials have assessed the initial use of room air versus 100% oxygen in the resuscitation of newborns.^9–14 Although the bulk of enrolled patients were of term gestation, infants as small as 1,000 g birth weight and 27 weeks gestation were included in 4 of the trials.^9–11,¹⁴ In general, these studies indicated that using room air as a starting point during resuscitation resulted in similar outcomes to 100% oxygen. Most of the investigations found that infants resuscitated with room air took less time to have the onset of both their first breath and their first cry. Approximately one quarter of the room air-treated neonates needed to be switched to 100% oxygen after several minutes because of a failure to respond. There have been many criticisms of these investigations. The 4 trials done in developing countries^9–11,¹⁴ were quasi-randomized, they enrolled patients into the treatment groups depending on whether it was an odd-number versus even-number day, and the investigators were not blinded to the F_IO2 children received. Two of the trials^10,12 found trends toward higher mortality in the 100% oxygen groups. The other 4 studies, including the 2 performed in a developed country,^12,13 did not find a similar trend. Two trials revealed increases in markers of oxidative stress (eg, serum levels of glutathione and superoxide dismutase) among those treated with 100% oxygen. Short-term neurological outcomes (assessed at 3–28 days) were reportedly equivalent in all 6 investigations. The 4 studies that enrolled preterm infants did not stratify outcomes by gestational age.

Sorely lacking in all of these reports were outcome data concerning major morbidity: chronic lung disease, persistent pulmonary hypertension, intracranial hemorrhage, cerebral white matter injury, necrotizing enterocolitis, patent ductus arteriosus, and retinopathy of prematurity. There was only one long-term follow-up study, of the infants in the largest trial.¹⁵ Approximately one third of infants in that trial were subsequently assessed in an unblinded fashion at 18–24 months following enrollment. The examiners were investigators from the original study and were not neurologists. They administered a questionnaire and examined the children. Cognition was not assessed. There were no differences in the outcomes between the groups in that limited follow-up.

The data generated by these studies have been controversial. Interestingly, a group of independent authors, using the aforementioned data, published meta-analyses in 2 separate reports published within months of each other and drew opposite conclusions.^16,17 In their initial publication, a Cochrane Systematic Review,¹⁶ they concluded that the evidence was insufficient to recommend room air over 100% oxygen. Several months later, using the same data and analytical methods, they concluded that room air should initially be used for resuscitating term and near-term infants, with oxygen as a backup.¹⁷ Despite the limitations of the 6 investigations, the International Liaison Committee on Resuscitation recently revised its recommendations concerning the use of oxygen: “In term infants receiving resuscitation at birth with positive-pressure ventilation, it is best to begin with air rather than 100% oxygen. If, despite effective ventilation, there is no increase in heart rate, or if oxygenation (guided by oximetry) remains unacceptable, use of a higher concentration of oxygen should be considered.”¹⁸ This was stated to be a class IIb, level of evidence B recommendation,¹⁹ meaning that it is an acceptable recommendation based on “weak” evidence (not excellent or good evidence). In the recently published 6th edition of the Textbook of Neonatal Resuscitation, published under the auspices of The Neonatal Resuscitation Program, appears a table of preductal saturation target ranges during resuscitation (Table 1).^19,20 If an infant's oxygen saturation is below the lower limit at specified time points, it is recommended that oxygen should be administered, via blender, in concentration sufficient to achieve the target range, and that use of room air is advocated (but not mandated) at the initiation of resuscitation in a term infant.²⁰

What should one do concerning supplemental oxygen when resuscitating a premature infant? In the 4 aforementioned studies, the premature infants' outcomes were not delineated. The optimal oxygen saturation for the resuscitation, stabilization, and ongoing care of the very preterm infant remains undefined.²¹ There have been several small, randomized controlled trials of lower versus higher F_IO2 during neonatal resuscitation of preterm infants.^22–27 Only one of the trials used room air as their initial low-F_IO2 group,²² and every one of those infants needed supplemental oxygen within 2–5 min of birth. Similar to the term-gestation infant trials, there appeared to be fewer markers of oxidative stress among infants treated with lower F_IO2.^23,25,27 None of those trials were adequately powered to evaluate important long-term outcomes such as survival, neurodevelopmental disability, or BPD. For premature infants the International Liaison Committee on Resuscitation recommends: “Because many preterm infants of < 32 weeks gestation will not reach target saturation on room air, blended oxygen and air may be given judiciously, and is ideally guided by pulse oximetry. Both hyperoxemia and hypoxemia should be avoided.”¹⁸ The new 6th edition of the Textbook of Neonatal Resuscitation advocates “something less than 100% oxygen, but more than 21% oxygen for a preterm baby.”²⁰ In my unit, when resuscitating a premature infant we start with an initial F_IO2 of 0.50 and adjust upward or downward depending on the preductal oxygen saturation during the resuscitation process. A large well designed randomized controlled blinded prospective trial is needed to determine the appropriate F_IO2 when initiating resuscitation.

In his classic tome Retrolental Fibroplasia: A Modern Parable,²⁸ the late William Silverman, the “grandfather” of randomized controlled trials in neonates, related a story of oxygen restriction in preterm infants. During the 1950s it was recognized that premature infants given high F_IO2 during hospitalization were more likely to develop the serious, potentially blinding condition of retrolental fibroplasia (ie, retinopathy of prematurity). There was subsequent widespread restriction of oxygen use in premature infants. Only after the ensuing 5 years was it recognized that this oxygen restriction had the unintended consequence of increasing mortality and cerebral palsy in survivors. Additionally, several large trials are underway to compare 2 different target oxygen saturation ranges (85–89% and 91–95%) in premature infants < 28 weeks gestation, and Stenson²⁹ recently examined mortality data from more than 3,600 infants in those trials. He found a significant 8.5% higher frequency of death among infants in the lower target saturation range, and the difference appeared to be mainly due to more deaths from brain injury and necrotizing enterocolitis. I would caution clinicians to be judicious about jumping on the low-F_IO2 or room air bandwagon.

CPAP and Sufficient PEEP

The lungs of preterm infants at birth are surfactant deficient, so it may be difficult for them to achieve adequate functional residual capacity (FRC) and maintain open alveoli and distal airways. The delivery room use of CPAP or PEEP of at least 4–5 cm H₂O has been advocated to assist in lung expansion, establish FRC, and improve oxygenation.^1,3,4 Early use of CPAP in the delivery room, in both preterm and term-gestation infants, has been enthusiastically advocated by many clinicians,²⁰ but there is a scarcity of data supporting this approach.

Finer et al³⁰ assessed the effects of CPAP or PEEP in infants < 28 weeks gestation. One-hundred four infants were randomized to either no CPAP and 100% oxygen or to CPAP/PEEP via a T-piece resuscitator (NeoPuff, Fisher & Paykel, Auckland, New Zealand). Delivery room intubation was needed in 41% of the controls and 49% of the CPAP/PEEP group, whereas there was no significant difference in intubation at some other time during hospitalization (78% vs 82%). Death was more likely in the CPAP/PEEP group (27% vs 13%, P = .07), as was the incidence of pneumothoraces (13% vs 9%, difference not significant). The frequency of chronic lung disease was not assessed.

Morley and colleagues³¹ randomized more than 600 premature infants (25–28 weeks gestation) in the delivery room to CPAP (8 cm H₂O) or to intubation and ventilation. Although the CPAP group had a trend toward a lower combined end point of death or BPD (34% vs 39%), the difference was not significant. In the CPAP group there was less use of surfactant, and the subjects averaged one day less of mechanical ventilation. Unfortunately, the CPAP-treated infants were significantly more likely to develop pneumothoraces (9% vs 3%, P < .001).

The SUPPORT trial³² involved infants born between 24 and 27 weeks gestation. The infants were randomized to either intubation and surfactant within 1 hour of birth, or to CPAP (5 cm H₂O) initiated in the delivery room. The CPAP patients could subsequently be intubated and receive surfactant if they fulfilled certain criteria (F_IO2 > 0.50, P_aCO2 > 65 mm Hg, or hemodynamic instability). There was a combined primary outcome: death or BPD. Eighty-three percent of the CPAP group subsequently required intubation. The CPAP group had approximately 3 fewer days of mechanical ventilation. Nevertheless, there were no statistically significant differences in death, BPD, or the combined end point of death or BPD.

Members of the Vermont Oxford Network³³ performed a prospective randomized controlled trial of 3 approaches in the delivery room. Preterm infants (n = 648) of 26–29 weeks gestation were randomized to either: prophylactic surfactant followed by ventilation; intubation and surfactant followed by rapid extubation to nasal CPAP; or early stabilization on nasal CPAP. The nasal CPAP group was less likely to receive surfactant or to be ventilated during the first week of life. However, there were no differences between groups in the outcomes of death, chronic lung disease, or other complications of prematurity.

Finally, the Brazilian Network on Neonatal Research³⁴ also assessed early use of nasal CPAP in the delivery room. They randomized 173 infants (birth weight 1,000–1,499 g) to either nasal CPAP (5 cm H₂O) immediately in the delivery room, or to nasal CPAP subsequently when clinically indicated. Early use of nasal CPAP in the delivery room did not improve the frequency of surfactant use, ventilator use, pneumothoraces, or BPD.

I would conclude from the aforementioned trials that the current evidence to support routine use of CPAP/PEEP in the delivery room among preterm infants is less than compelling. If we examine the major end points of death or BPD, the use of CPAP/PEEP in the delivery room does not influence outcomes. Finally, there are virtually no scientific data on the delivery room use of CPAP or PEEP in term-gestation infants. The International Liaison Committee on Resuscitation report¹⁸ and the Neonatal Resuscitation Program guidelines²⁰ suggest that the use of CPAP/PEEP should reflect local expertise and preferences.

Surfactant

Exogenous surfactant therapy has been an integral part of neonatal intensive care unit care for the past 20 years. In the preterm infant population, surfactant use has substantially decreased mortality. However, it has not affected BPD incidence. Premature lung disease (respiratory distress syndrome) results from surfactant deficiency in a structurally immature lung that is easily injured.⁴¹ The major surfactants used throughout the world in 2011 are animal-derived, generally from a mince or lavage of either bovine or porcine lungs.⁴²

Outside the preterm population, there are several disorders in which surfactant deficiency or inactivation occurs (eg, meconium aspiration syndrome, pneumonia). Several trials have found exogenous surfactant effective in non-premature infants. In a systematic review, Soll and colleagues⁴³ concluded that prophylactic use (generally defined as within the first 15 min of life) of exogenous surfactant in premature infants is better than selective (rescue) use of surfactant among infants < 29 weeks gestation (fewer pneumothoraces, decreased mortality, less pulmonary interstitial emphysema). Prophylaxis is generally done in the delivery room soon after birth. Most of the data supporting the superiority of prophylaxis come from studies performed in the late 1980s and early 1990s, which was before the widespread use of antenatal corticosteroids (to hasten lung maturation) and newer methods of support (eg, high-frequency ventilation) and modes of conventional ventilation (eg, synchronization, volume-targeting, pressure support). Additionally, many believe that CPAP at a sufficient level accomplishes results similar to exogenous surfactant (by opening the alveoli and improving oxygenation). Thus, frequently clinicians are now delaying surfactant administration until they are certain respiratory distress syndrome is established and is more than mildly debilitating.

The surfactant approach used in the Intubation, Surfactant, Extubation to CPAP (INSURE) trial has been widely practiced in Europe over the past decade⁴⁴ and has gained popularity in other areas of the world. Although most randomized controlled trials of the INSURE approach have found a short-term decrease in the need for mechanical ventilation, no differences in mortality or chronic lung disease have been described.⁴⁴ The Vermont Oxford Network trial³³ found no advantages with the INSURE approach. The current place of INSURE therapy during resuscitation needs to be clarified.

Do we need new surfactants? With currently available animal-derived surfactants, up to 15% of premature infants fail to respond. In the commercially available surfactants there is wide variability (± 50%) of target surfactant protein level, and potential immunogenicity and infection risks. Synthetic surfactant that contains surfactant protein mimics may be one way of improving our current therapies.⁴² In addition, noninvasive delivery (eg, aerosolization or via laryngeal mask airway) may prove useful in certain situations.

Although surfactant therapy is used daily in delivery rooms around the world in premature infants, to date neither the International Liaison Committee on Resuscitation nor the Neonatal Resuscitation Program have addressed surfactant use in the delivery room. Thus, clinicians have to make their own decisions on the timing of surfactant therapy in the vulnerable population of preterm infants.

Oropharyngeal and Intratracheal Suctioning of Meconium-Stained Amniotic Fluid

Prevention of meconium aspiration syndrome⁷ has long been a goal of clinicians in the delivery room. Approximately 12–15% of all newborn infants are born through meconium-stained amniotic fluid, and, historically, 3–10% of those patients develop meconium aspiration syndrome.^7,45,46 It is associated with substantial morbidity, such as need for mechanical ventilation, persistent pulmonary hypertension of the newborn, need for extracorporeal membrane oxygenation, long-term abnormal pulmonary dysfunction, and death. Histological features of meconium aspiration syndrome include inflammation, obstruction of airways, and areas of atelectasis alternating with overexpansion. Meconium can inactivate endogenous surfactant and can cause direct injury to the lung parenchyma, including necrosis and hemorrhage. Multiple maneuvers have been used in the delivery room to prevent meconium aspiration^7,45,46 The hope has been that removal of meconium-stained amniotic fluid would mitigate its effects on the infant's lungs.

For a 25-year period starting in the mid-1970s, the routine teaching was to intubate all meconium-stained infants and suction out the trachea.⁷ We performed a large randomized controlled trial to assess whether this maneuver would be effective in apparently vigorous meconium-stained neonates.⁴⁵ We enrolled more than 2,000 infants. If they were thought to be vigorous immediately after delivery (heart rate > 100 beats/min, good respiratory effort, normal tone), they were randomized to either intubation and tracheal suctioning or to expectant management. We found no differences in outcomes between the 2 groups. Additionally, since the mid-1970s it was routine when delivering a meconium-stained infant to suction out the nasopharynx and oropharynx before the head and shoulders are delivered.⁷ The thinking was that removing the meconium-stained fluid before the child's first breath would prevent aspiration. We performed a large, randomized controlled trial to assess whether this therapy was truly effective.⁴⁶ We enrolled more than 2,000 meconium-stained infants who were randomized to either intrapartum oropharyngeal and nasopharyngeal suctioning or to expectant management. There were no differences in outcomes between the 2 groups. These results have contributed to the thinking that meconium aspiration syndrome is often a result of in utero aspiration of meconium-stained amniotic fluid and cannot be prevented by intrapartum or postnatal therapies. There remains a major question of whether the depressed meconium-stained infant benefits from intratracheal suctioning, as has long been recommended by the International Liaison Committee on Resuscitation and the Neonatal Resuscitation Program. Over the past decade there have been many discussions about performing a definitive clinical trial, but at present there are no trials underway to answer this question.

Summary

Newly born infants frequently need assistance to establish normal respirations, but many resuscitation maneuvers can injure the lungs of preterm and term-gestation neonates. I have discussed the rationale for preventing such injury and the evidence basis for what we do. To date, however, there are no good scientific data supporting changes in our current delivery room management of neonates that can clearly increase lung protection and improve overall outcomes.