Patient-ventilator interaction in ARDS patients with extremely low compliance undergoing ECMO : a novel approach based on diaphragm electrical activity

Patients with acute respiratory distress syndrome (ARDS) requiring extracorporeal membrane oxygenation (ECMO) usually present very low respiratory system compliance (Cst(rs)) values (i.e., severe restrictive respiratory syndrome patients). As a consequence, they are at high risk of experiencing poor patient-ventilator interaction during assisted breathing. We hypothesized that monitoring of diaphragm electrical activity (EAdi) may enhance asynchrony assessment and that neurally adjusted ventilatory assist (NAVA) may reduce asynchrony, especially in more severely restricted patients. Methods: We enrolled ten consecutive ARDS patients with very low Cstrs values undergoing ECMO after switching from controlled to pressure support ventilation (PSV). We randomly tested (30 min) while recording EAdi: (1) PSV30 (PSV with an expiratory trigger at 30 % of flow peak value); (2) PSV1 (PSV with expiratory trigger at 1 %); (3) NAVA. During each step, we measured the EAdi-based asynchrony index (AIEAdi) = flow-, pressure- and EAdi-based asynchrony events/EAdi-based respiratory rate × 100. Results: AI EAdi was high during all ventilation modes, and the most represented asynchrony pattern was specific for this population (i.e., premature cycling). NAVA was associated with significantly decreased, although suboptimal, AI EAdi values in comparison to PSV30 and PSV1 (p < 0.01 for both). The PSV30-NAVA and PSV1-NAVA differences in AIEAdi values were inversely correlated with patients' Cstrs (R 2 = 0.545, p = 0.01 and R 2 = 0.425, p < 0.05; respectively). Conclusions: EAdi allows accurate analysis of asynchrony patterns and magnitude in ARDS patients with very low Cstrs undergoing ECMO. In these patients, NAVA is associated with reduced asynchrony.