Acute respiratory failure (ARF) is a life-threatening emergency which occurs due to impaired gas exchange. In the US, the number of hospitalisations owing to acute respiratory failure was 1,917,910 in the year 2009.(1) The incidence of ARF requiring hospitalization was 137.1 per 100,000 population.(2) In ARF due to chronic obstructive pulmonary disease (COPD) and cardiogenic pulmonary edema, non-invasive ventilation (NIV) has been shown to be beneficial. NIV also has several advantages over invasive mechanical ventilation. These include, avoidance of endotracheal intubation and its attendant complications like airway injury, nosocomial infections, and possibly shorter duration of intensive care unit (ICU) stay.(3, 4) The success of NIV depends on several factors like the etiology of the respiratory failure, careful monitoring by the treating physician, and also adequate cooperation of patient. Better synchrony of the patient's spontaneous breaths with the ventilator-delivered breaths may lead to better patient cooperation and thereby, better clinical outcomes. Patient-ventilator asynchrony (PVA) leads to dyspnea, increased work of breathing, and prolonged duration of mechanical ventilation.(5) Pressure support ventilation (PSV) is one of the commonest mode used during NIV. In a prospective multicenter observational study, severe asynchrony (defined as an asynchrony index of >10 %) was seen in 43% of patients of patients with ARF ventilated by NIV with the conventional PSV mode.(6) Neurally adjusted ventilator assist (NAVA) is new mode of ventilation which utilizes the electrical activity of the diaphragm to deliver the breath.(7) During NAVA, breath is delivered when the patient's diaphragm starts contracting. Further, the amount of pressure support given during the breath is proportional to the strength of the electrical signal from the diaphragm. Finally, NAVA also terminates the breath when the electrical activity of the diaphragm wanes. NAVA has been shown to avoid over-assistance, decrease intrinsic positive end-expiratory pressure (PEEP), and minimize wasted efforts.(8) Hence, NAVA may play a major role in improving patient-ventilator synchrony.

In a pooled analysis of studies comparing NAVA with PSV during NIV, it was shown that the use of NAVA significantly improved patient-ventilator synchrony.(9) However, so far, no clinical trial has demonstrated that this improvement in synchrony translates into better clinical outcomes. In this randomized controlled clinical trial, we intend to compare the rates of NIV failure and mortality between NAVA and PSV in subjects with acute respiratory failure managed with NIV.

NIV is mode of ventilation wherein positive pressure ventilation is delivered by non-invasive interface like nasal mask, face mask, nasal plugs or helmet avoiding invasive interface by endotracheal intubation or by tracheostomy. In 1980s NIV by mask was used with success in patients of obstructive sleep apnea and later on in neuromuscular respiratory failure. Use of NIV has increased subsequently over next 20 years with overall first line NIV being used upto 23% (10) NIV is definitely useful as first line of management in COPD exacerbation, cardiogenic pulmonary edema, post extubation respiratory failure and is also worth trial with conflicting data in cases of acute hypoxemic respiratory failure due to pneumonia, ARDS.

In cases of acute exacerbation of COPD, NIV improves clinical outcomes.(11-13) In a metaanalysis by Ram FS et al which included 14 RCT and 758 patients which showed decreased mortality(11 vs 21%), intubation rate(16 vs 33%) and treatment failure(20 vs 42%)(12). Similarly, there is high quality evidence supporting use in cardiogenic pulmonary edema showing to improve respiratory parameters and decrease intubation rates (14-17). A meta-analysis of 32 studies including 2916 patients in 2013 by Vital FM et al showed that NIV reduces in hospital mortality compared to standard medical care(RR 0.66, 95% CI 0.48 - 0.89)(14) NAVA as mode of ventilation was developed after initial landmark study "neural control of mechanical ventilation in respiratory failure" was published in 1998 by Sinderby(7). It was later on introduced on the servo-i ventilator in 2007. NAVA is a kind of pressure assist ventilation using electrical activity of diaphragm and thus the neural output to initiate the breath, regulate the assist level and cycle off the breath.

In study by Piquilloud et al of 13 patients which compared NAVA with PSV during NIV there was significant difference in asynchronies. Trigger delay (Td) was reduced with NAVA to 35 ms (IQR 31-53 ms) versus 181 ms (122 - 208 ms). Also, there were no premature or delayed cycling, ineffective efforts in the NAVA group. Asynchrony index (AI) with NAVA was 4.9% (2.2 - 10.5%) compared to 15.8% (5.5 - 49.6%) with PSV (20)In a recent meta-analysis comparing PSV with NAVA during NIV by Inderpaul et al which included 9 studies with 96 subjects including both adult and paediatric patients showed that asynchronies were more in PSV than in NAVA in both adult and paediatric studies. The overall pooled mean difference of asynchrony index was 28.02(95% CI, 11.61 - 44.42). also the risk of severe asynchrony was 3.4 % times higher in PSV compared to NAVA group.(9) NAVA mode of ventilation in previous studies has shown to decrease asynchrony with ventilator, mainly reducing ineffective efforts, cycling delays. Ineffective efforts may occur due to presence of intrinsic PEEP which in turn is more common during prolonged insufflations and at high levels of assist. Ineffective efforts may also occur with weak inspiratory efforts which may occur during states of high respiratory drive. NAVA uses electrical activity of diaphragm to trigger a breath and there is marked difference in improving synchrony with studies reporting no wasted efforts when compared to PSV mode. In a study by Vignaux et al showed that during conventional NIV, 40 % of the patients experience asynchrony and the rate of asynchrony correlated with leakage(6). In NAVA mode of ventilation assistance is delivered based upon the neural trigger and hence is not affected by leaks.

In a pooled analysis of studies comparing NAVA with PSV during NIV, it was shown that the use of NAVA significantly improved patient-ventilator synchrony.(9) However, so far, no clinical trial has demonstrated that this improvement in synchrony translates into better clinical outcomes. In this randomized controlled clinical trial, we intend to compare the rates of NIV failure and mortality between NAVA and PSV in subjects with acute respiratory failure managed with NIV.

Inclusion Criteria:

1. Respiratory rate >30 per minute
2. Arterial blood gas analysis showing a PaCO2 >45 mmHg and pH <7.35
3. PaO2/FiO2 ratio < 300
4. Use of accessory muscles of respiration or paradoxical respiration

Exclusion Criteria:

1. Age <18 years or >75 years
2. Pregnancy
3. PaO2/FiO2 ratio ≤100
4. Hypotension (systolic blood pressure <90 mmHg)
5. Severe impairment of consciousness (Glasgow coma scale score <8)
6. Inability to clear respiratory secretions (Airway care score [ACS] ≥12)(27)
7. Abnormalities that preclude proper fit of the NIV interface (agitated or uncooperative patient, facial trauma or burns, facial surgery, or facial anatomical abnormality)
8. Subjects who have an artificial airway like tracheostomy tube or T-tube
9. Contraindications for insertion of naso-/orogastric feeding tube (facial/nasal trauma, recent upper airway surgery, esophageal surgery, esophageal varices, upper gastrointestinal bleeding)
10. More than two organ failures
11. Unwillingness to undergo placement of nasogastric catheter
12. Known phrenic nerve lesions
13. Suspected diaphragmatic weakness
14. Patient already on home NIV therapy for chronic respiratory failure
15. Application of NIV for more than one hour for the current illness
16. Failure to provide informed consent

Complete protocol and individual patient data after publication of results