Background: Neurally adjusted ventilatory assist (NAVA) is a mode of ventilation controlled by the electrical activity of the diaphragm (Edi). The aim was to evaluate patient-ventilator interaction in infants during NAVA as compared with conventional ventilation.
Methods: Infants were successively ventilated with NAVA, pressure control ventilation (PCV), and pressure support ventilation (PSV). Edi and ventilator pressure (Pvent) waveforms were compared and their variability was assessed by coefficients of variation.
Results: Ten patients (mean age 4.3 ± 2.4 mo and weight 5.9 ± 2.2 kg) were studied. In PCV and PSV, 4 ± 4.6% and 6.5 ± 7.7% of the neural efforts failed to trigger the ventilator. This did not occur during NAVA. Trigger delays were shorter with NAVA as compared with PCV and PSV (93 ± 20 ms vs. 193 ± 87 ms and 135 ± 29 ms). During PCV and PSV, the ventilator cycled off before the end of neural inspiration in 12 ± 13% and 21 ± 19% of the breaths (0 ± 0% during NAVA). During PCV and PSV, 24 ± 11% and 25 ± 9% of the neural breath cycle was asynchronous with the ventilator as compared with 11 ± 3% with NAVA. A large variability was observed for Edi in all modes, which was transmitted into Pvent during NAVA (coefficient of variation: 24 ± 8%) and not in PCV (coefficient of variation 2 ± 1%) or PSV (2 ± 2%).
Conclusion: NAVA improves patient-ventilator interaction and delivers adequate ventilation with variable pressure in infants.