Nasal Intermittent Positive Pressure Ventilation for Neonatal Respiratory Distress Syndrome

: Nasal or noninvaisve intermittent positive pressure ventilation (NIPPV) refers to well-established noninvasive respiratory support strategies combining a continuous distending pressure with intermittent pressure increases. Uncertainty remains regarding the benefits provided by the various devices and techniques used to generate NIPPV. Our included meta-analyses of trials comparing NIPPV with continuous positive airway pressure (CPAP) in preterm infants demonstrate that both primary and postextubation NIPPV are superior to CPAP to prevent respiratory failure leading to additional ventilatory support. This short-term benefit is associated with a reduction in bronchopulmonary dysplasia, but not with mortality. Benefits are greatest when ventilator-generated, synchronized NIPPV is used.


INTRODUCTION
In the past, preterm infants with signs of moderate or severe respiratory distress were intubated and mechanically ventilated. This invasive approach resulted in inflammation of the lungs in the short-term and impaired development and scarring known as bronchopulmonary dysplasia (BPD) in the long-term. 1 Efforts to decrease rates of BPD in the surfactant/antenatal steroid era have led to an increased use of noninvasive respiratory support for even the most immature infants. 2 Prophylactic nasal continuous positive airway pressure (CPAP), started soon after birth, is now recommended for spontaneously breathing very preterm or very lowbirth-weight infants with respiratory distress syndrome (RDS). 3 Prophylactic CPAP reduces the need for mechanical ventilation and surfactant administration, and lowers the rates of both BPD alone and the combined outcome of death or BPD when compared with immediate endotracheal ventilation. 3 Despite the physiologic and clinical benefits, CPAP failure rates remain at approximately 50% in the first week of life in extremely preterm newborns at highest risk for developing BPD. [4][5][6][7] CPAP failure is associated with a substantial increase in important adverse outcomes including air leak, BPD, intraventricular hemorrhage, and death. 8 As a result, methods to augment the effectiveness of CPAP have gained interest. 9 Noninvasive intermittent positive pressure ventilation (NIPPV) applied at the nose has become a well-established therapy for preterm infants. 10 Despite its frequent use, uncertainty remains regarding the precise terminology, the appropriate clinical indications, the different devices and techniques used to generate NIPPV, and the level of benefit they provide. These differences complicate the interpretation of the available evidence. 11,12 In this review, we address these uncertainties, with a particular focus on NIPPV as primary and postextubation respiratory support for preterm infants with RDS. For both indications, we summarize the current evidence from randomized controlled trials (RCTs) comparing NIPPV with CPAP.

NASAL INTERMITTENT POSITIVE PRESSURE VENTILATION Terminology
NIPPV is a form of noninvasive respiratory support that combines a continuous positive end-expiratory airway pressure (PEEP) with intermittent higher pressures delivered by a nasal mask or nasal prongs. The terminology surrounding NIPPV is confusing. There are many alternative pressure generating devices, interfaces, and settings available. In addition, some devices allow synchronization with the infant's own breathing efforts. The most common abbreviations include nsNIPPV (nonsynchronized NIPPV), sNIPPV (synchronized NIPPV), BiPAP (biphasic CPAP), bilevel CPAP, and bilevel NIPPV.
Although all these modes are considered forms of NIPPV, 2 main NIPPV modalities must be distinguished: traditional NIPPV, with settings designed to mimic ventilator settings, typically generated using a ventilator, or alternatively settings that are more reflective of bilevel CPAP, typically generated using flow-drivers. There are devices of both types which have the capacity to synchronize pressure changes with spontaneous breathing ( Table 1).

Pressure and Volume Delivery
The lower pressure level (PEEP) during NIPPV offers the same physiologic benefits as CPAP, that is, stabilization of the upper airways and the compliant preterm chest wall and prevention of end-expiratory alveolar collapse. This maintains functional residual capacity and reduces ventilation-perfusion mismatch, which improves oxygenation and work of breathing. The intermittent pressure peaks increase the mean airway pressure (MAP) above the PEEP level, potentially further recruiting the lung, which may improve functional residual capacity more efficiently than CPAP alone. 13 Table 1 Characteristics of 'traditional' NIPPV and bilevel CPAP 18 Because of large and variable leaks around the nose and mouth, the transmission of applied NIPPV pressures to the lung is substantially attenuated. 14,15 Moreover, observational studies demonstrate that the majority of nonsynchronized pressure peaks occur during spontaneous expiration and do not contribute to tidal volume. 16 When the pressure rises coincided with spontaneous inspiration, only a 15% increase in relative tidal volume was noted. During apneic episodes, pressure peaks resulted in measurable tidal volumes only 5% of the time, and produced tidal volumes a quarter of those seen during spontaneous breathing. Higher peak inspiratory pressures did not increase the likelihood of a visible chest inflation, suggesting that higher set pressures may not provide additional respiratory assistance during apnoea. 16 Whether nonsynchronized NIPPV confers any benefit over CPAP when the PEEP during CPAP is matched to the generated MAP during NIPPV is still a matter of debate. A small crossover study including 10 infants on nonsynchronized NIPPV and CPAP delivered at the same MAP found minimal differences in short-term outcomes, suggesting that any advantage of nonsynchronized NIPPV may arise from a higher MAP rather than from the effect of the intermittent pressure peaks themselves. 13

Synchronization
Observations of low pressure and volume delivery during nonsynchronized NIPPV suggest support may be more effective if inflations are synchronized with the infant's own inspiratory efforts. Synchronization may be achieved by airway flow detection, which ensures that the glottis is open before pressure is applied. 17 However, this is challenging because of air leakage around the prongs and masks and from the open mouth. Graseby capsules are unaffected by air leak, but may be affected by movement artifact; however, they are the most commonly used method for NIPPV synchronization. 18 These cheap, lightweight, and disposable capsules are noninvasively attached to the anterior abdominal wall below the xiphoid process; they consist of a small, flat balloon filled with air, which is sensitive to pressure variations. The balloon connects to a pressure transducer capable of detecting the beginning of the diaphragmatic contraction, which enables the synchronization of the pressure peak. Although the accuracy of the Graseby capsule is affected by its position, method of fixation, and movement artifacts, it produces reliable signals that rapidly trigger the set pressure peak with most spontaneous breaths. [19][20][21] Other potential synchronization methods include neurally adjusted ventilatory assist, currently available with the Servo-n ventilator (Maquet, Solna, Sweden) and respiratory inductance plethysmography. 22

Safety
Although there were initial concerns regarding an increased risk of gastrointestinal side effects with NIPPV, recent evidence suggests that NIPPV is a safe therapy in preterm infants. 23 This has been confirmed by 2 systematic reviews of the Cochrane Collaboration on NIPPV for initial support of neonatal RDS and for preterm infants after extubation. 11,12 Both reviews reported no significant differences between the NIPPV and CPAP groups in rates of feeding intolerance, gastrointestinal perforation, necrotizing enterocolitis, or air leak. The incidence of nasal injury through tight-fitting binasal prongs has not been assessed systematically for infants receiving NIPPV. Since the risk of nasal injury, and the strategies to prevent it are considered the same for NIPPV and CPAP, use of nasal masks, rotating nasal interfaces, and nasal barrier dressings may be equally effective in reducing nasal injury during NIPPV. 24

CLINICAL EVIDENCE
The majority of clinical trials in preterm infants have compared NIPPV with CPAP as either the primary mode of treatment for neonatal RDS, or after extubation. Of these trials, Kirpalani's NIPPV Trial dominates the literature. 25 This large, pragmatic trial differs from the smaller studies in that it recruited a heterogeneous study population and permitted a variety of devices to deliver NIPPV, including some that delivered synchronized pressure changes. Although pragmatic, the considerable degree of methodological and clinical heterogeneity makes interpretation of pooled trial results difficult. To evaluate the impact of these variations, we begin with a review of Kirpalani's NIPPV Trial and its substudies, and provide updated meta-analyses of trials comparing NIPPV with CPAP as primary or postextubation support for neonatal RDS.

Kirpalani's Nasal Intermittent Positive Pressure Ventilation Trial
This large randomized, controlled, multicenter trial conducted between 2007 and 2011 hypothesized that NIPPV would reduce the risk of BPD in extremely low-birth-weight infants by minimizing the duration of endotracheal intubation. 25 Infants with a birth weight of less than 1000 g and a gestational age of less than 30 weeks, eligible for noninvasive support within the first 28 days of life, were randomly assigned to 1 of 2 forms of noninvasive respiratory support, NIPPV or CPAP. Initial settings for respiratory support were provided, but not mandated and clinicians could individualize care. No NIPPV delivery devices were specified, NIPPV synchronization was permitted but not mandated. The primary outcome was a composite of death or moderate/severe BPD according to National Institutes of Health criteria. 26 Three preplanned subgroup analyses were performed according to birth weight, prior intubation status (intubated or nonintubated before randomization), and the form of the intervention used in the NIPPV group (synchronized or nonsynchronized).
A total of 1009 infants with a mean gestational age of 26 weeks and a mean birth weight of 800 g were enrolled. The primary outcome, death or BPD occurred in 38.4% (191 of 497 infants) randomized to NIPPV and in 36.7% (180 of 490) randomized to CPAP (adjusted odds ratio, 1.09; 95% confidence interval [CI], 0.83-1.43; P 5 .56). There were no significant differences between NIPPV and CPAP in the individual components of death or BPD, in other prespecified secondary outcomes including potential adverse effects of treatment, or in the subgroup analyses according to birth weight, prior intubation status, or synchronization.
In the years following the publication of Kirpalani's NIPPV Trial results, 2 secondary analyses have been published with the following aims: (1) to examine whether important outcomes differed in infants who received ventilator-generated or flow-driver-generated NIPPV, and (2) to compare noninvasive ventilation failure rates in intubation-naïve extremely low-birth-weight infants randomized to NIPPV or CPAP. 27,28 Substudy 1: ventilator-generated versus flow-driver-generated nasal intermittent positive pressure ventilation This nonrandomized comparison from Kirpalani's NIPPV Trial provides outcome data on the 497 infants in the NIPPV group. 27 NIPPV could be delivered by a ventilator or a flow-driver device based on unit preference, practice, and device availability. Irrespective of the device, traditional NIPPV settings or bilevel CPAP settings could be used. In the NIPPV group, 215 infants received ventilator-generated NIPPV and 241 received flow-driver-generated NIPPV. Forty-one infants, in whom both devices had been used, were excluded. The composite outcome, death or BPD at 36 weeks was 39% in the ventilator-generated NIPPV group and 37% in the flow-driver-generated NIPPV group (adjusted odds ratio, 0.88; 95% CI, 0.57-1.35; P 5 .56). Although rates of BPD Nasal Intermittent Positive Pressure Ventilation were not significantly different between groups (adjusted odds ratio, 0.64; 95% CI, 0.41-1.02; P 5 .061), more deaths occurred before 36 weeks gestational age in the flow-driver-generated NIPPV group (2.3% vs 9.4%; adjusted odds ratio, 5.01; 95% CI, 1.74-14.4; P 5 .003).

Substudy 2: nasal intermittent positive pressure ventilation versus continuous positive airway pressure in intubation-naïve infants
The second substudy compared the rate of 'failure of noninvasive support' in infants who were never intubated before enrollment and randomization. 28 As opposed to the original trial and substudy 1, the primary outcome was defined as failure of noninvasive respiratory support requiring endotracheal intubation at any time in the first 7 days after randomization. Of the 1009 extremely low-birth-weight infants initially enrolled in the NIPPV trial, 142 had not been intubated before randomization. Of those, 27.5% in the NIPPV group and 30.1% in the CPAP group were subsequently intubated (relative risk, 0.91; 95% CI, 0.54-1.53). The combined outcome of death or BPD at 36 weeks postmenstrual age was not different between groups (19.7% vs 16.7%; risk ratio, 1.18; 95% CI, 0.58-2.40). There was no significant difference in rates of air leak.
What do the results of Kirpalani's nasal intermittent positive pressure ventilation trial and its substudies mean?
In contrast with the results obtained from the pooled analysis of smaller trials that favored the use of NIPPV in preterm infants, Kirpalani's NIPPV trial and its substudies found no significant benefit of NIPPV with respect to the risk of death or survival without BPD. 11,12,[29][30][31] There may be several reasons for the differences in findings. First, more immature infants were included in Kirpalani's NIPPV trial (Tables 2 and 3); failure of noninvasive support is more prevalent in extremely preterm infants and is associated with a marked increase in the rate of adverse outcomes, including death and BPD. 4,8 In such a high-risk population with RDS due to surfactant-deficient lungs, collapsing airways and poor muscle strength, a number of infants may still be inadequately supported with NIPPV despite the modest increase in MAP provided by additional positive pressure breaths. Second, the pragmatic trial design did not specify the ventilator device, settings, or use of synchronization. In the NIPPV group, approximately half of infants received flow-driver-generated NIPPV, typically set to deliver modest peak pressures, lower than pressures set during ventilatorgenerated NIPPV. Indeed, mortality was higher in infants who mostly received flowdriver-generated NIPPV, possibly due to a higher reintubation rate compared with infants receiving ventilator-generated NIPPV (adjusted rate ratio for number of reintubations, 1.23; 95% CI, 1.02-1.49). 27 Moreover, a subgroup analysis by synchronization rather than by device revealed that ventilator-generated NIPPV was mostly applied in a nonsynchronized manner, and synchronization was more often used during flow-driver-generated NIPPV (suggesting that traditional NIPPV settings with short high-pressure durations were still commonly used during flow-driver-generated NIPPV, cf. Table 1). Both combinations of device and technique may be associated with a lack of effective pressure transmission to the lungs, and may contribute to the finding of no significant benefit of NIPPV.

Meta-Analyses of Trials Comparing Nasal Intermittent Positive Pressure Ventilation with Continuous Positive Airway Pressure
The updated meta-analyses were performed using RevMan, version 5.4. 32 Relevant studies were identified by searching PubMed, The Cochrane Library, and the reference lists of included articles. Studies were included if they were RCTs that enrolled  preterm infants (born <37 weeks' gestation), compared any form of NIPPV with CPAP, and reported the primary outcome. Four subgroup analyses included whether NIPPV was delivered by a ventilator or by a flow-driver-device and whether pressure changes were synchronized with spontaneous breathing or not. No differentiation was made between the type of settings (traditional NIPPV vs bilevel CPAP) applied in each trial.

Rü egger et al
A fixed-effect model was used to pool data of included trials.

Primary respiratory support
This meta-analysis aimed to compare the efficacy of NIPPV versus CPAP when used as primary respiratory support in preterm infants with RDS who were less than 6 hours old. We defined the primary outcome as respiratory failure leading to additional ventilatory support during the first week of life. An early and brief period of endotracheal ventilation for an INSURE (intubate-surfactant administration-extubate) procedure was not considered as respiratory failure. We pooled data from 18 trials and 1900 infants (see Table 2). We added data from 8 newly published trials comprising 850 infants to the existing meta-analysis of the Cochrane Collaboration. 11 In 11 trials, ventilator-generated NIPPV with variable peak pressures of 12 to 26 cm H 2 O were used. Five trials used flow-driver-generated support with peak pressures of 8 to 15 cm H 2 O. Two trials, including Kirpalani's NIPPV trial used mixed methods: both ventilator-generated and flow-driver-generated NIPPV, with or without synchronization, and with variable pressure settings. None of the included studies attempted to match the PEEP in the CPAP group with the generated MAP in the NIPPV group.
Six individual trials reported a significant reduction in rates of respiratory failure during the first week of life in infants managed with NIPPV. Twelve showed no significant difference, and none showed a significant benefit for CPAP. Pooled data from all 18 trials demonstrated a clinically important, 37% relative reduction in the risk of respiratory failure with NIPPV (Fig. 1). This beneficial effect was most obvious in the trials using ventilator-generated NIPPV (combined subgroups 1.1.1 and 1.1.2: RR 0.60; 95% CI, 0.48-0.76), and was greatest when synchronization was used (subgroup 1.1.2), with 7 infants needing to be treated with ventilator-generated, synchronized NIPPV to prevent one respiratory failure. Flow-driver devices provided a smaller (subgroup 1.

Postextubation respiratory support
This meta-analysis evaluated the efficacy of NIPPV versus CPAP when used as postextubation respiratory support for preterm infants. We defined the primary outcome as respiratory failure leading to additional ventilatory support during the week postextubation.
Fifteen trials enrolling 2444 infants were included (see Table 3). Data from 5 trials published between 2016 and 2020 enrolling 1013 infants were added to the existing meta-analysis of the Cochrane Collaboration. 12 Twelve trials used ventilatorgenerated NIPPV, 2 trials delivered flow-driver-generated NIPPV, the final study being the Kirpalani mixed methods study.
Overall, infants extubated to NIPPV had a 22% relative risk reduction for respiratory failure within the first week postextubation compared with those managed with CPAP (Fig. 2). Extubation of 12 infants to NIPPV would prevent one case of extubation failure. Again, this benefit was greatest in trials using ventilator-generated NIPPV (combined subgroups 1.2.1 and 1.2.2: RR 0.51; 95% CI, 0.40-0.65) and was strongest when ventilator-generated, synchronized NIPPV was used (subgroup 1.2.2). On average, only 3 infants would need to be extubated to ventilator-generated, synchronized NIPPV to prevent one case of postextubation failure. No beneficial effect was found for flow-driver devices (subgroup 1.2.3).

Mortality
Results from 17 trials enrolling 1834 infants could be pooled for this analysis. Overall and within subgroups, no difference in mortality was noted when NIPPV was compared with CPAP as primary respiratory support for preterm infants with neonatal RDS ( Table 4). After extubation, no significant reduction in mortality was found in the meta-analysis of 10 trials and 2178 infants, and no significant difference was detected when mortality was examined by device or synchronization for either primary or postextubation support.
These findings contrast the results of the respective meta-analysis of the Cochrane Collaboration on NIPPV versus CPAP for preterm neonates after extubation, where a small difference in mortality between treatment groups was reported, favoring NIPPV (RR 0.69; 95% CI, 0.48-0.99). 12

Bronchopulmonary dysplasia
Fourteen of the 18 trials (1534 infants) evaluating NIPPV versus CPAP as primary respiratory support reported BPD at 36 weeks' corrected gestational age. We noted a 28% relative reduction in the risk of BPD with NIPPV ( Table 5). However, this overall difference was fully attributable to the reduction in BPD seen in studies using ventilator-generated, synchronized NIPPV. None of the other subanalyses showed a significant difference in the rate of BPD between groups. In the meta-analysis of the Cochrane Collaboration on NIPPV as primary respiratory support for preterm infants, no reduction in BPD was observed in any of the subgroups. 11 After extubation, meta-analysis of 11 studies enrolling 2128 infants revealed a borderline lower rate of BPD when infants were randomized to NIPPV compared with infants randomized to CPAP (see Table 5). Once again, within subgroups, only ventilator-generated, synchronized NIPPV was associated with a reduction in BPD.

SUMMARY
There is clear evidence that NIPPV is superior to CPAP as primary and postextubation respiratory support for the prevention of respiratory failure in preterm infants with RDS. For both indications, ventilator-generated, synchronized NIPPV is most effective to prevent respiratory failure. Results show no reduction in mortality overall or within subgroups, irrespective of whether primary or postextubation NIPPV support is delivered. Longer-term pulmonary benefits include a reduction in BPD, but only with ventilatorgenerated, synchronized NIPPV. Implementation of this evidence may be hampered by the limited availability of ventilators able to deliver synchronized pressure changes. There is little evidence of harm during NIPPV generated by any device or delivered in any mode, with no reported increase in abdominal adverse events.
In view of the high heterogeneity among trials included in our meta-analyses, results may not be generalizable and must be interpreted with caution. It is important for clinicians to understand that not all modes of noninvasive support are the same, and variations in the applied strategy may not provide the same level of benefit. Superiority of ventilator-generated NIPPV over flow-driver-generated NIPPV is explained by higher peak pressures used during ventilator-generated NIPPV. If MAPs were matched across all devices and all modes, there may be little difference between CPAP, flow-driver-generated NIPPV, and ventilator-generated NIPPV.
Additional data from adequately powered RCTs are warranted to determine the benefits of NIPPV in smaller and more immature infants. A particular focus should be placed on clinically relevant outcomes such as death, BPD, and long-term respiratory function following prolonged NIPPV use. Moreover, the role of synchronized NIPPV as primary respiratory support starting directly after birth in the delivery room deserves further attention.

BEST PRACTICES
NIPPV is preferable over CPAP as primary and post-extubation respiratory support in preterm infants with RDS.
For both indications, ventilator-generated, synchronized NIPPV should be used to prevent respiratory failure.
Ventilator-generated, synchronized NIPPV may reduce the risk of bronchopulmonary dysplasia when used as either primary or post-extubation support in preterm infants, but is not associated with a decrease in mortality.

CLINICS CARE POINTS
Not all modes of NIPPV are the same, and variations in the applied strategy may affect the level of benefit.
During nonsynchronized NIPPV, most pressure peaks occur during spontaneous expiration and do not contribute to tidal volume.
Any advantage of nonsynchronized NIPPV may arise from a higher mean airway pressure rather than from the effect of the intermittent pressure peaks themselves.
Synchronization of the positive pressure peaks with the infant's own breathing efforts results in a more effective pressure and volume delivery. Nasal Intermittent Positive Pressure Ventilation