2019 Year in Review: Neonatal Respiratory Support

Suctioning Infants With Meconium-Stained Amniotic Fluid

Approximately 10–15% of all infants are born in the presence of meconium-stained amniotic fluid.² Of these, 3–9% go on to develop meconium aspiration syndrome.² It has been known for some time that routine intubation and tracheal suctioning of all neonates with meconium-stained amniotic fluid does not result in a decreased incidence of meconium aspiration syndrome.³ Because there is insufficient evidence to support the routine practice of tracheal suctioning in nonvigorous infants (defined as having depressed respiration and poor muscle tone) born with meconium-stained amniotic fluid, suctioning is not recommended.² Positive-pressure ventilation is recommended by the American Heart Association and the American Academy of Pediatrics for nonvigorous infants born with meconium stained amniotic fluid, but only if the infant is apneic or the heart rate fell to < 100 beats/min.⁴ Edwards et al⁵ conducted a retrospective database review of 301,150 infants from United States centers participating in the Vermont Oxford Network admitted to the neonatal ICU before the guideline change (2013–2015) and after the change (2017). Overall, neonatal ICU admissions for meconium aspiration syndrome decreased from 1.8% to 1.5% of total admissions (relative risk [RR] 0.82, 95% CI 0.68–0.97). In the delivery room, endotracheal tube suctioning decreased from 57.0% to 28.9% from before the guideline change to the current era (RR 0.51, 95% CI 0.41–0.62). Adjunct therapies, including high-frequency ventilation, inhaled nitric oxide (INO), and extracorporeal membrane oxygenation, were unchanged between the groups. The authors note that the use of surfactant in patients with meconium aspiration syndrome increased from 24.6% to 30.0% (RR 0.82, 95% CI 1.01–1.48) in 2017. Although the authors state that mortality increased from 2.6% to 2.9% (RR 1.12, 95% CI .74–1.69) and moderate to severe hypoxic-ischemic encephalopathy increased from 5.4% to 6.8% (RR 1.24, 95% CI 0.91–1.69). Importantly, the confidence intervals are wide, and it is unknown what interventions or clinical features may have confounded this retrospective study. Indeed, the current guidelines do not recommend routine suctioning, but clinicians may feel inclined to offer suctioning in certain cases. Edwards et al⁵ were unable to comment about this rate because further work is needed before a concrete conclusion can be made.

Kumar et al⁶ conducted a randomized controlled trial (RCT) of endotracheal suctioning for the prevention of meconium aspiration syndrome. They enrolled 132 nonvigorous neonates with meconium-stained fluid to receive either endotracheal suctioning or no suctioning. No differences were observed in the incidence of meconium aspiration syndrome (31.8% vs 22.7%, RR 1.4, 95% CI 0.793–2.470), mortality (13.6% vs 7.5%, P > .05), or duration of hospital stay (54 vs 44 h, P > .05) in the endotracheal suctioning or no suctioning groups. Routine suctioning in the moments following birth with meconium-stained amniotic fluid was not found to be useful in preventing meconium aspiration syndrome.

Supplemental Oxygen

Over the last quarter century, the overall survival in extremely premature babies has improved.^7,8 Because more infants are surviving, there is an increased incidence of intraventricular hemorrhage, necrotizing enterocolitis, patent ductus arteriosus, developmental delay, bronchopulmonary dysplasia (BPD), and retinopathy of prematurity.⁸ Supplemental oxygen and mechanical ventilation are important risk factors for the development of BPD and retinopathy of prematurity.^9,10 Certainly, supplemental oxygen is an essential component of the overall life-sustaining care that is provided to premature babies as part of resuscitation, but are there ways to facilitate pulmonary vasodilation and support adequate oxygenation without high levels of oxygen? Sekar et al¹¹ investigated the use of INO as an additional therapy during neonatal resuscitation with a double-blind, RCT of 28 infants during positive-pressure ventilation. Half were randomized to receive oxygen ( = 0.3) and INO at 20 ppm (the INO group), and the other half were randomized to receive oxygen and placebo (the control group). The authors noted that cumulative oxygen exposure was significantly lower in the INO group compared to the control group (P = .001). The proportion of hyperoxia was also lower in the INO group compared to the control group (P < .001). There were no differences in the measured physiologic parameters, nor was there a need for invasive mechanical ventilation. Although use of INO is understood to be safe when used at levels ≤ 20 ppm, it is not clear from the current data that adding INO to reduce has an important effect on the risk of BPD and retinopathy of prematurity. However, it will be interesting to see this work grow in the coming years as more work will be done.

A number of large RCTs have sought to determine the effect of low (85–89%) and high (91–95%) targets in extremely premature infants.^12–14 Lower saturation ranges are associated with an increased risk of death but with a reduced risk of retinopathy of prematurity.^12–14 As such, the optimal saturation range continues to be debated. Foglia et al¹⁵ sought to investigate the association between oxygen saturation alarm policy changes and neonatal morbidity and mortality. They conducted a retrospective review of 3,809 subjects in 10 hospitals with a change in alarm policy and 3,685 neonates from 9 hospitals with no changes in their policy between 2006 and 2014. In hospitals with a policy change, differences in morbidity and mortality were compared in the period before and after the policy change. The policy change primarily included tightening the upper and lower alarm limits. Mortality was similar in hospitals that did have a policy change over the study period (adjusted odds ratio 0.94, 95% CI 0.75–1.18), but it decreased in hospitals who did not institute a policy change (adjusted odds ratio 0.63, 95% CI 0.5–0.8). Stated more simply, hospitals that enacted tight alarm targets did not see an improvement in mortality, and those that ignored this initiative tended to see a decrease in mortality. Although these data are retrospective, the findings suggest that tight alarm limits are not associated with reduced mortality or incidence of retinopathy of prematurity, as one may suspect.

Comparing High-Flow Nasal Cannula and Nasal CPAP

Recently, Guimarães et al¹⁸ conducted a retrospective analysis of 135 premature infants whose birthweight was < 1,500 g. Nasal trauma was reported in 65% of subjects receiving CPAP, 26% of whom had skin breakdown of grade 2 or worse. Therefore, many groups have sought alternative treatment modalities. Heated, humidified, high-flow nasal cannula (HFNC) therapy has exhibited efficacy and safety similar to CPAP when applied as a primary approach to mild and moderate respiratory distress in premature babies and may reduce the likelihood of nasal trauma and pneumothorax.¹⁹ Hong et al²⁰ conducted a meta-analysis of RCTs comparing the effectiveness and side effects of HFNC and CPAP that included data from 21 RCTs and 2,886 premature infants. Overall, treatment failure was similar between HFNC and CPAP (RR 1.03, 95% CI 0.79–1.33). HFNC was associated with a reduced risk of nasal trauma (P < .001) and pneumothorax (P = .03) versus CPAP. In the case of providing respiratory support following extubation, CPAP was associated with a lower chance of treatment failure (RR 1.23, 95% CI 1.01–1.50). In babies with primary respiratory support, there were no differences in the time to reach full feeds between HFNC and CPAP. On the surface, this finding is perplexing because one would think that increased comfort and reduced abdominal distention during HFNC would allow a quicker advance on feedings. However, many trials do not protocolize the feeding regimen when comparing respiratory support devices, and the similarities in time to full feeds may be due to practice. Trials specifically designed to assess the advancement and tolerance of full feeds between respiratory support modalities are warranted.

Physiologic Effects of HFNC

Overall, HFNC continues to be a safe and mostly effective tool to provide respiratory support to premature infants. The mechanisms of action are not well understood, however, although they are thought to include a reduction in dead-space, generation of end-expiratory pressure, and possibly even agitation.²¹ During HFNC, a number of factors influence the relative effectiveness of the therapy. Degree of leak around the nares, mouth open or closed, degree of spontaneous breathing, and flow settings have important effects on the degree of pressure generation and therefore on support.^21,22

Liew et al²³ conducted a prospective randomized crossover study to assess the effects of HFNC on respiratory physiology. Among 44 infants whose birthweight was < 2,000 g, increasing flows from 2 to 8 L/min led to a mean increase in end-expiratory pressure of 2.3–6.1 cm H₂O as measured with nasopharyngeal airway pressure monitoring. However, the variance among subjects and in subjects when the mouth was closed or open was great. In some cases, the end-expiratory pressure exceeded 12 cm H₂O. In Figure 1, the variance in nasopharyngeal pressure as a function of set flow on the HFNC is depicted.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Scatter plot of relationship between nasopharyngeal end-expiratory pressure and weight-adjusted flow. There is large variability of end-expiratory pressure measured > 6 L/min/kg, with some measurements measuring up to 8–13 cm H₂O. From Reference 23, with permission.

Charles et al²⁴ conducted a prospective, randomized, crossover study of HFNC and NIV to compare the work of breathing for premature infants following extubation. To assess work of breathing, catheters were placed to measure esophageal and gastric pressures, and the transdiaphragmatic pressure was calculated to obtain an estimate of the work of breathing. By virtue of the crossover design, each subject served as their own control and received 2 h of either HFNC or NIV followed by 2 h of the opposing therapy in a randomized fashion. Although 21 subjects were enrolled, only 9 completed the study and were therefore included in the analysis. The pressure-time product of the transdiaphragmatic pressure was lower during NIV than HFNC (232 vs 365 cm H₂O × s/min; P = .008). Applying these findings to other patient populations is tricky. First, the NIV was set to deliver a peak pressure of 14–16 cm H₂O and an expiratory pressure of 5 cm H₂O. Because there is no increase in support during the inspiratory phase of the breath cycle with HFNC, it is not surprising that NIV provides more support and therefore decreases the total pressure-time product of the transdiaphragmatic pressure. Although the authors concluded that the lower estimated work of breathing during NIV is advantageous, that may not always be the case. Indeed, the essential question to answer is what the target work of breathing for a premature infant in the ICU actually is. Certainly, in the context of acute decompensation and limited respiratory effort, NIV may be indicated, but if a patient is supported without distress on HFNC, why would one select a therapy that offers a reduced work of breathing? This study offers a reasonable comparison of NIV and HFNC, but larger RCTs are required to determine which is most effective.

INSURE Technique

The INtubation-SURfactant-Extubate (INSURE) technique is one of the most commonly used methods to provide SRT to preterm infants with RDS. The INSURE technique requires a brief exposure to invasive mechanical ventilation with the aim of extubation to noninvasive support; however, a subset of infants cannot be extubated successfully, many of whom were previously stable on CPAP.^16,30,60 De Bisschop and colleagues⁶¹ performed a systematic review with the aim of identifying early predictive risk factors for INSURE failure in preterm subjects with RDS. These authors found a median INSURE failure rate of 33% (9.3–52%) and identified ELBW (ie, 750–1,000 g), lower gestational age, and severe RDS as risk factors for failure. Due the methodological heterogeneity of included studies, it was not possible to create a clinical tool to distinguish infants who could be successfully extubated after the INSURE technique from those who may benefit from continued mechanical ventilation. Similarly, Awaysheh et al⁶² described a failure rate of 25% in subjects < 30 weeks gestational age who were treated with the INSURE technique and mechanical ventilation for < 2 h. They reported that gestational age < 28 weeks, birthweight < 1,000 g, and pH < 7.0 were risk factors for INSURE failure.

Novel Delivery Approaches

The high failure rate of the INSURE technique and perceived benefits of avoiding intubation and mechanical ventilation altogether have led to the development of less invasive methods of SRT, including pharyngeal or laryngeal delivery, inhalation, and thin catheter administration (TCA), which is synonymously described as minimally invasive surfactant therapy or less invasive surfactant administration. Of these methods, TCA via feeding tube, flexible angiocatheter, or specialty introducer tubes has been most extensively investigated.⁵⁹ The TCA procedure requires direct or video-assisted laryngoscopy, followed by the insertion of a catheter into the trachea and through the vocal cords of spontaneously breathing infants.⁶³ Although methods vary slightly, surfactant can be delivered while the infant continues to receive noninvasive support. A recent meta-analysis of RCTs directly comparing TCA and INSURE have reported a reduction in BPD or death composite (RR 0.66, 95% CI 0.46–0.93) and a decrease in the need for mechanical ventilation in the first 72 h (RR 0.72, 95% CI 0.53–0.97), with marginal or no difference in pneumothorax and mortality, respectively.⁵⁹ These finding are consistent with previous meta-analyses that reported improved survival without adverse events for infants treated with TCA.^64–68 These data contribute to the growing body of evidence supporting the benefits of less invasive surfactant administration, which has been adopted in Europe as the preferred method of SRT for spontaneously breathing infants.⁶⁹ It is important to note that TCA is more commonly practiced in Europe and Australia. In the United States, only 15% of surveyed neonatologists reported using TCA (8% as routine care and 7% for research).⁷⁰ Additionally, the procedure requires training, and there are several areas that require further investigation, including the establishment of specific treatment thresholds according to gestational age, timing of treatment, and whether analgesia or sedation is required.⁶³

Nebulized Surfactant

Inhaled surfactant represents the least invasive method of SRT and has received considerable research interest. However, due to technical difficulties, including the high viscosity of surfactant, the ability to reproducibly aerosolize particles 1–5 µm in diameter without protein denaturation, and the uncertainty of optimal concentration or dosing, have hampered adoption into clinical practice.^59,71 A recent, single-center, blinded RCT stratified subjects by 29–31 or 32–33 weeks gestational age and compared the efficacy of CPAP alone to CPAP with nebulized surfactant, with the primary outcomes being intubation and the requirement of mechanical ventilation at 72 h.⁷² Surfactant nebulization reduced the requirement for intubation compared with CPAP alone (RR 0.53, 95% CI 0.29–0.95). However, this reduction was limited to the 32–33 weeks gestational age stratum, and there were no differences in the proportion of subjects who remained intubated after 24 h, 72 h, or 7 d. This was the first study to provide evidence of successful noninvasive SRT using the eFlow vibrating mesh nebulizer (Pari, Starnberg, Germany). In vitro evaluations of the eFlow nebulizer have reported favorable aerosol characteristics (mass median particle diameter = 3 µm).⁷³ These authors also evaluated dose-response in spontaneously breathing rabbits with severe respiratory distress and noted that surfactant doses of 200–400 mg/kg significantly improved oxygenation, respiratory indices, and compliance, achieving a similar response to animals treated with an intratracheal bolus of 200 mg/kg. This nebulization strategy is currently being evaluated in a multi-center phase 2 clinical trial including preterm subjects with mild to moderate RDS (https://www.clinicaltrialsregister.eu/ctr-search/search?query=2016-004547-36, Accessed February 7, 2020).

Predicted Need for Invasive Ventilation

In a secondary analysis of RCT data, Roberts and colleageues^74,75 sought to identify predictors of both early intubation and short duration respiratory support (≤ 1 d) in subjects born at 28–36 weeks gestational age. Subjects < 30 weeks gestational age with ≥ 0.3 were independent predictors of intubation within 72 h with a corresponding likelihood ratio of 9.1. Conversely, subjects ≥ 34 weeks getstational age with = 0.21 were 5 times more likely to require only brief respiratory support. These data corroborate the findings of previous studies evaluating NIV failure and suggest that a subset of older infants > 34 weeks gestational age may receive noninvasive support unnecessarily.

Volume-Targeted Ventilation

The greatest need for intubation and mechanical ventilation is observed in the smallest and most premature infants. The use of volume-targeted ventilation over pressure-limited ventilation confers a reduction of BPD, hypocapnia, grade III–IV intraventricular hemorrhage, pneumothorax, and ventilator duration.⁷⁶ Despite high-level evidence in support of volume-targeted ventilation, acceptance into clinical practice has occurred slowly because this represents a major paradigm shift away from decades of mostly pressure-limited ventilation use. Several barriers have hampered the diffusion of volume-targeted ventilation, including provider experience level, lack of established V_T targets based upon disease state, large endotracheal tube leaks (eg, > 50%), and a ventilator's ability to deliver and accurately measure V_T.^77–79

Wong and colleagues⁸⁰ prospectively studied very low birthweight and ELBW subjects who underwent volume-targeted ventilation with the aim of describing breath-to-breath variation in expired V_T. The authors analyzed 6 h of continuous ventilator data and found no significant variations in expired V_T between cohorts, despite the ELBW cohort having significantly lower set V_T (4.6 vs 7.2 mL, P < .001). This study evaluated volume-targeted ventilation delivery in a single ventilator type because previous bench and animal studies have suggested considerable variability in ventilator performance, with variations in V_T becoming more pronounced with increased leak or lower V_T targets.^79,81 Future studies should evaluate V_T target ranges for specific disease conditions, such as evolving RDS, congenital diaphragmatic hernia, meconium aspiration syndrome, and BPD, and incorporate performance data respective to different ventilators.

High-Frequency Ventilation

A subset of infants are managed with high-frequency ventilation (HFV) as the primary mode of ventilation. In many cases, HFV is implemented primarily as a rescue modality in babies who exhibit persistent respiratory failure despite maximum conventional lung-protective ventilation strategies. Overall, studies comparing the use of HFOV and high-frequency jet ventilation to conventional mechanical ventilation have not reported significant improvement in outcomes for elective or rescue usage.^82,83 A recent, single-center, retrospective cohort of actively managed extremely premature infants were separated in 2 groups: 22–23 gestational age and 24–25 weeks gestational age.⁸⁴ The investigators assessed survival and neurological outcomes at 18–22 months of corrected age. Both groups received antenatal steroids, intubation, SRT, and high-frequency jet ventilation as the primary mode of ventilation. Survival to discharge for subjects 22–23 weeks gestational age was 78% (95% CI 0.69–0.88) compared to 89% (95% CI 0.84–0.93) for subjects 24–25 weeks gestational age. In surviving infants, no or mild neurologic impairment was present in 64% (95% CI 0.50–0.77) and 76% (95% CI 0.68–0.83) for the 22–23 and 24–25 weeks groups, respectively. Although the study was conducted over a 10-y period, the results suggest favorable survival and neurodevelopmental outcomes for infants who are often perceived as being on the fringe of viability. In addition, extremely premature infants were not included in previous RCTs evaluating primary use of HFV. The role of primary HFV in extremely premature infants warrants further prospective investigation.

The Rescue-HFOV Trial Group prospectively evaluated factors affecting the response of 372 subjects who underwent rescue HFOV after failing conventional ventilation.⁸⁵ Subjects were separated into 2 groups: survivors and non-survivors or those who received extracorporeal membrane oxygenation (n = 8). All subjects initially underwent volume-controlled ventilation with V_T < 7 mL/kg and were managed with a permissive hypercapnia strategy to maintain a pH > 7.25. HFOV was initiated when the oxygen requirement exceeded 0.6 to achieve > 90%. Overall survival to discharge was 58%; subjects who died had lower birthweight (1,655 ± 1,091 g vs 1,858 ± 1,027 g, P < .006), higher initial (0.83 vs 0.72, P < .001), and higher rate of INO exposure (29% vs 11%, P = .004). Gestational age was similar between groups; however, in a subgroup analysis of lower gestational age was associated with non survivors (28.6 ± 3.6 vs 29.8 ± 3.6 weeks, P = .006). The Rescue-HFOV study included a large multi-center cohort using an open-lung approach during HFOV and used several ventilators that are not commercially available in the United States. These data afford a unique perspective because extracorporeal membrane oxygenation was only available in 1 out of 23 sites, which underscores the importance of early identification of HFOV failure.