Comparison of neurally-adjusted ventilator assist in infants before and after extubation

To compare invasive (iNAVA) and non-invasive (nivNAVA) neurally adjusted ventilatory assist in infants, respect to gas exchange, breathing pattern, respiratory drive, infant-ventilator interaction and synchrony, vital parameters and required sedation.

Ten consecutive intubated term infants admitted for respiratory failure of different etiology underwent to 2-hour not-randomized trials of iNAVA and, after extubation, nivNAVA, the latter with unchanged ventilator settings and with air-leaks compensating software. Arterialized capillary blood was sampled at the end of each trial. We computed: 1) the minimum (EAdimin) and peak (EAdipeak) values of the diaphragm electrical activity; 2) ventilator (RRmec) and own patients' (RRneu) respiratory rates; 3) inspiratory (delayTR-insp) and expiratory trigger delays (delayTR-exp) and the time of synchrony between patient's effort and ventilator assistance (Timesynch/Tineu); 4) the asynchrony index. Vital parameters and required sedation were also recorded.

iNAVA and nivNAVA did not differ between in terms of gas exchange (pH (7.35 [7.31-7.41] vs. 7.36 [7.30-7.40], P=0.745), PcCO2 (38.4 [34.8-42.6] vs. 36.9 [33.9-41.6] mmHg, P=0.469) and PcO2/FiO2 (211 [168-323] vs. 214 [189-282], P=0.195), respectively). EAdimin, EAdipeak, RRmec and RRneu were similar before and after extubation. Both modes confirmed an optimal infant-ventilator interaction (i.e. delayTR-insp, delayTR-exp and Timesynch/Tineu), irrespective of the interface, and no patients showed clinical relevant asynchronies. A low requirement of sedation with fentanyl was recorded during both trials, without differences between.

We found iNAVA and nivNAVA to be characterized by similar gas exchange, breathing pattern, respiratory drive, infant-ventilator interaction and synchrony, vital parameters and required sedation.