Neurally adjusted ventilator assist (NAVA) reduces asynchrony during non-invasive ventilation for severe bronchiolitis.

Abstract

To determine the prevalence of main inspiratory asynchrony events during non-invasive intermittent positive-pressure ventilation (NIV) for severe bronchiolitis. Ventilator response time and asynchrony were compared in neurally adjusted ventilator assist (NAVA) and in pressure assist/control (PAC) modes. This prospective physiological study was performed in a university hospital's paediatric intensive care unit and included 11 children (aged 35.2 ± 23 days) with respiratory syncytial virus bronchiolitis with failure of nCPAP. Patients received NIV for 2 hr in PAC mode followed by 2 hr in NAVA mode. Electrical activity of the diaphragm and pressure curves were recorded for 10 min. Trigger delay, main asynchronies (auto-triggering, double triggering, or non-triggered breaths) were analyzed, and the asynchrony index was calculated for each period. The asynchrony index was lower during NAVA than during PAC (3 ± 3% vs. 38 ± 21%, P < 0.0001), and the trigger delay was shorter (43.9 ± 7.2 vs. 116.0 ± 38.9 ms, P < 0.0001). Ineffective efforts were significantly less frequent in NAVA mode (0.54 ± 1.5 vs. 21.8 ± 16.5 events/min, P = 0.01). Patient respiratory rates were similar, but the ventilator rate was higher in NAVA than in PAC mode (59.5 ± 17.9 vs. 49.8 ± 8.5/min, P = 0.03). The TcPCO2 baselines values (64 ± 12 mmHg vs. 62 ± 9 mmHg during NAVA, P = 0.30) were the same and their evolution over the 2 hr study period (-6 ± 10 mmHg vs. -12 ± 17 mmHg during NAVA, P = 0.36) did not differ. Patient-ventilator inspiratory asynchronies and trigger delay were dramatically lower in NAVA mode than in PAC mode during NIV in infants with severe bronchiolitis.