The aim of this study is to demonstrate a significant decrease in asynchrony with NIV-NAVA using the Servo n ventilator (Getinge, Sweden), as compared to abdominal triggered (Graseby capsule) synchronized nasal intermittent positive pressure ventilation (sNIPPV) using the Infant Flow CPAP device (Care Fusion, USA).

All of the data obtained can be used to develop a large-scale study aimed at reducing the rate of re-intubation in the study population (pilot study). In fact, the re-intubation criteria for extremely premature children are based on clinical criteria (desaturations, apnea, signs of respiratory control) and paraclinical criteria (FiO2, Potential hydrogen (pH), PCO2).

The results of this pilot study will help to develop an adapted methodology and to calculate a sample size to compare the 2 modes of NIV to the test on a clinical criterion: the rate of re-intubation after extubation, which is classically high in these patients.

The use of non-invasive ventilation has significantly reduced morbidity and mortality in premature newborns by reducing the pulmonary lesions caused by invasive ventilation. Currently, variable flow continuous positive airway pressure (CPAP) devices, such as the infant flow® driver, are considered more efficient than constant flow pressure sources. Nasal intermittent positive pressure ventilation, as compared to CPAP, might reduce the extubation failure rate, but has no impact on mortality or bronchopulmonary dysplasia. However, data is lacking on the interest of synchronization and on the effect of the different available interfaces (prongs, masks, cannulas). In addition, the ventilatory characteristics (high respiratory rate and low inspiratory effort) of the premature infant increase the risk of asynchrony between the patient and the ventilator, which is a major cause of poor tolerance for this type of ventilation.

NAVA (neurally adjusted ventilatory assist) is a recent ventilatory mode that offers proportional assistance to respiratory work based on the measured electrical activity of the diaphragm via oesophageal electrodes. It thus allows a regulation of inspiratory pressures and time by the patient him/herself. The physiological effects of NAVA have been primarily described in intubated neonates and studies have shown a significantly improved synchronization and significantly decreased inspiratory pressures in patients ventilated with NAVA compared to intermittent controlled ventilatory support. However, the currently available evidence is limited and no beneficial effect on morbidity or mortality has been identified so far .

There are few studies on noninvasive NAVA (NIV-NAVA) conducted exclusively in neonates, most of which included a limited number of patients. Only one study to date compared NIV-NAVA to another synchronized NIV mode (NIV pressure support) using the Servo-i ventilator. This prospective crossover study found a significant decrease in peak inspiratory pressure (PIP), FiO2, frequency and length of desaturations in the NIV-NAVA group.

Decreased asynchrony has been observed during NIV-NAVA as compared to pressure-support NIV In adult patients and in 6 children hospitalized in the Pediatric ICU (median age 18 months).

In premature neonates, variable flow CPAP is preferentially used. Synchronized intermittent positive pressure can be delivered using a variable flow device and a Graseby abdominal capsule. Since variable flow CPAP is considered the most efficient pressure generator, it is legitimate to compare synchronization performance of the variable flow synchronized nasal intermittent positive pressure ventilation (sNIPPV) to NIV-NAVA. This comparison has never been performed so far, to our knowledge.

We hypothesize that synchronization will be markedly improved with NIV-NAVA as compared to sNIPPV.

Inclusion criteria:

• Premature infants born before 28 weeks of gestation
• Corrected age below 32 weeks of gestation
• Postnatal age > or = 3 days
• Receiving NIPPV (any mode)
• Equipped with an Edi catheter
• Receiving caffein treatment
• Parental consent
• Recipient of French social security coverage

Non-inclusion criteria:

• More than 1 apnea/hour requiring bag-mask ventilation, or pH<7.2 and/or TcPCO2>70, or FiO2>0.6 in the previous 6 hours.
• Nasal trauma precluding the use of non-invasive ventilation
• Major congenital anomalies
• Grade III or higher intraventricular hemorrhage
• Use of anesthetics or sedative within the past 24 hours, except opioids for iatrogenic withdrawal treatment
• Hemodynamic compromise defined as a mean blood pressure less than gestational age (in mmHg) or a capillary refill time more than 3 seconds
• Neuro-muscular disorders