In preterm infants, endotracheal mechanical ventilation is well known to cause various forms of lung injury including volutrauma, barotrauma and oxytrauma - collectively known as ventilator-induced lung injury (VILI). As such, there is a move towards non-invasive ventilation (NIV) in recent decades. However, many infants placed on NIV modes fail and require intubation. A relatively recent mode of NIV is non-invasive high frequency ventilation (NIHFV). Studies on this modality are scarce, but do suggest that neonates can be ventilated effectively. However, its efficacy in comparison with other existing modes of NIV remains unknown due to lack of appropriate studies. The investigators hypothesize that NIHFV is a superior NIV mode than Biphasic Nasal Continuous Positive Airway Pressure (BP-NCPAP) in preterm infants in preventing intubation following NCPAP failure.

Intubation and mechanical ventilation has long been the mainstay in managing infants with respiratory disease in the neonatal intensive care unit (NICU). However, in recent decades, the use of endotracheal mechanical ventilation (EMV) has been implicated as a major cause of ventilator-induced lung injury (VILI).Prolonged endotracheal intubation in the neonate has been linked with morbidities including bronchopulmonary dysplasia (BPD), nosocomial pneumonia, air leak syndromes and subglottic stenosis. Moreover, BPD has been linked with impaired neurodevelopmental outcome.

There has been a conscious effort amongst neonatologists to reduce the duration of EMV to minimize these undesirable side effects. Non-invasive ventilatory support has been increasingly used to that effect in recent years. Various forms of NIV including NCPAP, BP-NCPAP and nasal intermittent positive pressure ventilation (NIPPV) have become part of the ventilatory management of neonates in NICUs across the globe.

Originally described in the early 1970s,9 the use NCPAP has seen renewed interest since the 1990s when the significance of VILI became apparent. In retrospective studies, NCPAP has been shown to reduce the rates of bronchopulmonary dysplasia (BPD). A recent multi-center prospective randomized controlled trial of infants between 25 and 28 week gestation showed that infants treated with early NCPAP required fewer days of ventilation and had a reduced oxygen dependency after 28 days (although not at 36 weeks corrected GA). A subsequent analysis of these patients showed improved lung mechanics and gas exchange. However NCPAP use is not without its risks. Air leak syndromes, diminished cardiac output, gaseous distention and nasal injury/excoriations are well described complications of its use. Despite these drawbacks, NCPAP remains an efficacious modality to help reduce endotracheal intubation and limit VILI.

Although successful in reducing the reliance on EMV, it is also well known that up to 43-80% of neonates with moderate to severe respiratory disease treated with NCPAP fail and require intubation. BiPhasic NCPAP and NIPPV have been used with increasing frequency in NICUs in recent years to help prevent re-intubation, but with limited evidence guiding their use.

BiPhasic-NCPAP, described in the late 1980s as a form of augmented ventilation, provides two levels of non-invasive pressure support, typically cycled between 20-30 times per minute and the upper level of support usually lasting one second. As in NCPAP, the patient breathes around the two levels of pressures provided. However, there is currently no evidence to suggest that the use of BP-NCPAP offers any advantage over NCPAP. In a recent randomized controlled prospective study from the investigators centre comparing BP-NCPAP (N=69) and NCPAP (N=67), the former was not shown to be superior to the latter in facilitating sustained extubation in infants < 1,250 grams (67% vs. 58%, P = 0.27). Nevertheless, it is presumed that BP-NCPAP offers the same benefits as NCPAP, and remains in use as a rescue mode of NIV across many NICUs after NCPAP failure.

NIPPV is yet another NIV mode, offering a combination of NCPAP superimposed with ventilator provided peak inspiratory pressure (PIP) via a non-invasive interface, designed to mimic and augment the spontaneous breathing of neonates as in EMV. The PIP provided can be both synchronized or non-synchronized with patient breaths. However there is potential for significant inaccuracies in the synchronization process. Furthermore, the Infant Star ventilator with the Graseby capsule, one of the prototype ventilators capable of providing synchronized NIPPV, is no longer available. Despite this limitation, NIPPV remains in use in both non-synchronized and synchronized forms.8 At this time, to the best of the investigators knowledge, no studies have compared BP-NCPAP and NIPPV directly. In fact, in the recently completed NIPPV trial by Kirpalani et al BP-NCPAP and NIPPV were included in the same arm, which failed to show a reduction in the incidence of BPD compared to the NCPAP arm.

Based on studies conducted so far there is no firm consensus on what is the best non-invasive mode to be used and the modality of choice remains an individual decision depending on the preference and experience of the specific unit. Noninvasive high frequency ventilation is an alternative and relatively new mode of NIV and has been explored in 4 clinical studies so far (appendix A). In a study of 21 preterm and term infants, patients were switched from NCPAP to NIHFV to promote carbon dioxide (CO2) elimination and continued for a mean of 35 (2-144) hours. The authors reported that the partial pressure of carbon dioxide (PaCO2) decreased with the use of this modality from 8.3 to 7.2 kilo pascal (KPa). The major limitations of this study include the heterogeneity of patient population, the etiologies of primary lung disease, small sample size and lack of any control group. In a more recent study by Colaizy et al, NIHFV was used in 14 subjects for a period of 2 hours and a trend towards a decrease in PaCO2 levels was noted with the use of this modality. Limitations of this study include the short duration of NIHFV, small sample size and lack of a control group. The only randomized trial evaluating this modality included 40 term infants with clinically diagnosed transient tachypnea of newborn (TTN) and concluded that the use of NIHFV leads to quicker recovery of TTN compared with the use of CPAP. However the assessment of recovery was subjective, and in a study that was unblinded could be ridden with bias. As well, many of those patients may not have required any form of NIV in the first place. Lastly a study by Czernik et al evaluated a group of 20 neonates at high risk of post-extubation failure who were prophylactically placed on NIHFV. Fourteen of the infants were successfully weaned off to NCPAP, but there was no control group to compare these results with. None of the four trials reported any adverse effects related to the use of NIHFV, although they were not powered to evaluate such outcomes.

At two of the investigators centres (Mount Sinai Hospital and the Hospital for Sick Children) NIHFV has been used as a rescue mode in recent months when NCPAP and/or BP-NCPAP has failed. In a review conducted [unpublished at this time] more than half the instances of its use (out of the 21 total cases) were due to increased spells (apneas, bradycardias and/or desaturations). In the six hours preceding and the six hours after initiation of NIHFV, the number of spells reduced from an average of 5.2 (±3.5) to 1.5 (±1.3). The majority of patients tried on NIHFV remained extubated, with less than a third eventually requiring re-intubation. As of yet, no reports of harm were directly linked with the use of NIHFV.

Despite limited evidence, NIHFV holds promise as an alternate mode of NIV. It may be superior to NCPAP/BP-NCPAP with respect to PaCO2 clearance from the lungs and also has the advantage of not requiring breath-synchronization, an issue that has hindered the widespread use of NIPPV. However, definitive assessment of the efficacy of NIHFV requires a prospective, randomized controlled trial. The goal of this study is to compare the rates of intubation when using BP-NCPAP vs. NIHFV as rescue mode of NIV post NCPAP failure in infants < 1,250 grams.

Inclusion Criteria:

• Greater than 72 hours post-natal age AND currently on NCPAP
• Less than 1,250 g when on NCPAP at time of enrollment

Exclusion Criteria:

• Congenital or acquired abnormality of upper airways
• Severe congenital anomalies including cyanotic congenital heart disease
• Severe nasal excoriation/injury preventing use of NIV interface
• Greater than 2,000 grams at time of randomization