The similar rates of nasal septal injury, pneumothorax, and abdominal distention on bubble NIPPV in comparison with bubble NCPAP suggest that bubble NIPPV has a similar safety profile as bubble NCPAP for preterm infants in respiratory distress.
An important limitation of this study is the fact that patients were not randomized but were instead pragmatically allocated to bubble NIPPV vs. bubble NCPAP. Both groups had similar gestational ages and birth weights, but there was a significant difference in initial degree of respiratory distress. Specifically, the bubble NIPPV patients presented with worse respiratory distress on average as evidenced by higher mean pretreatment FiO2, Silverman Anderson Score and Respiratory Severity Score values in comparison with patients on bubble NCPAP. As noted previously, 11 of the patients in the bubble NIPPV group qualified for surfactant prior to initiation of bubble NIPPV; of these patients, 9 received surfactant within two hours of NICU admission [15]. Some of these patients had their first dose of surfactant administered prior to or concurrent with initiation of bubble NIPPV. In contrast, only 2 patients in the bubble NCPAP group qualified for surfactant prior to initiation of bubble NCPAP; both patients received surfactant within two hours of NICU admission.
After initiation of treatment, the clinical condition of patients in both the bubble NIPPV and NCPAP arms improved over the first six hours and subsequently stabilized. There was a greater improvement in Silverman Anderson Score and FiO2 in the bubble NIPPV group in comparison with the bubble NCPAP group. However, as previously noted, the bubble NIPPV group had more significant respiratory distress on average than the bubble NCPAP group prior to treatment initiation. We do not know whether the sicker infants would have had a similar response if they were instead placed on bubble NCPAP.
Excluding surfactant administered within two hours of admission to the NICU, there were similar rates of first and second dose of surfactant administration in both treatment arms. One patient was treated on NIPPV and then weaned to bubble NCPAP; the following day there was an acute change in respiratory status for which the patient was treated with surfactant. This was recorded as surfactant administered to the bubble NIPPV arm due to an intention to treat analysis.
As a field, neonatology is increasingly moving towards noninvasive therapies for supporting infants in respiratory distress. Noninvasive modalities are associated with lower rates of bronchopulmonary dysplasia and ventilator associated pneumonia in comparison with invasive modalities [16]. Additionally, the use of invasive modalities in resource constrained settings carries additional risk given high patient to nursing ratios and inconsistent availability of blood gas and patient monitoring equipment.
With regards to specific modalities of noninvasive ventilation of preterm infants, NCPAP is an effective treatment. For preterm infants in mild to moderate respiratory distress, NCPAP has been shown to reduce respiratory failure, need for mechanical ventilation and mortality in comparison with supplemental oxygen [17]. However, for infants in moderate to severe respiratory distress, NIPPV has been shown to provide additional benefit. Specifically, in preterm infants (28 + weeks gestational age) in respiratory distress, early initiation of NIPPV has been shown to further reduce respiratory failure necessitating intubation and mechanical ventilation in comparison with NCPAP [8]. NIPPV has also been shown to decrease post extubation failure in comparison with NCPAP [9]. Finally, there may be additional benefits to the use of NIPPV over NCPAP in infants with apnea [18].
This additional benefit for preterm infants in moderate to severe respiratory distress with NIPPV vs. NCPAP may be due to improvements in both oxygenation and ventilation. With regards to improved ventilation, the delta pressure (Phigh- Plow) can help facilitate improved CO2 clearance. With regards to improved oxygenation, a higher overall mean airway pressure can be delivered with NIPPV in comparison with NCPAP. For example, with a low pressure of 8 cm H2O, a high pressure of 20 cm H2O, a cycling rate of 40 cycles per minute and an inspiratory time of 0.5 seconds, a mean airway pressure of 12 cm H2O can be delivered. For comparison, many hospital protocols restrict the upper level of CPAP pressure to a lower level such as 8 to 10 cm H2O out of concern for air leaks.
The intermittent “breaths” in NIPPV can be delivered in synchrony with the patient's breaths (“SNIPPV") or in asynchrony (NIPPV). Historically all NIPPV was delivered asynchronously; anecdotally, patients appeared to synchronize their breathing efforts with the NIPPV. Overall, both NIPPV and SNIPPV appear to be equally beneficial [19]; to date there have not been randomized controlled trials comparing NIPPV with SNIPPV [20], although there is a study protocol published in 2017 [21].
Unfortunately, both SNIPPV and NIPPV have hitherto remained inaccessible for many resource constrained settings as equipment is expensive and requires continuous electricity. Donated ventilators used in a noninvasive mode can provide a short-term solution, but ongoing maintenance and procurement of replacement parts can be a challenge. To address this problem, a simple, non-electric bubble NIPPV technology was developed. While feasibility of this bubble NIPPV technology has been previously demonstrated, this study represents the first study of safety in the clinical setting of a NICU on a larger scale.
Overall, the similar rates of clinically relevant complications (nasal septal injury, pneumothorax, and abdominal distention) and secondary complications (NEC, IVH, PDA, ROP, sepsis) with a potential signal of improved physiologic efficacy (FiO2, Silverman Anderson Score) on bubble NIPPV in comparison with bubble NCPAP suggest that bubble NIPPV has a similar safety profile as bubble NCPAP for preterm infants in respiratory distress. Randomized controlled trials comparing the efficacy of bubble NIPPV with bubble NCPAP and ventilator derived NIPPV are warranted.