LAUSR

Introduction Respiratory mechanic is challenging for patients treated with veno-arterial Extracorporeal Membrane Oxygenation (VA ECMO). Whereas growing evidences were reported of changes in diaphragm thickness during mechanical ventilation (1,2), no study has reported an impact of VA ECMO settings on diaphragm effort. We hypothesized that ECMO settings influenced diaphragm thickening during ECMO weaning. We assessed diaphragm thickening fraction (TF) at the time of recovery from heart failure. Methods We conducted a nine-month prospective monocentric observational study. Adult patients ready for a first VA ECMO weaning trial, breathing spontaneously and hemodynamically stable at 2 l/min of VA ECMO flow were included. Sweep gas flow (SGF) was set for 15 min at 4 l/min, 2 l/min and 1 l/min with a 10 min return to baseline between each step (3). Mechanical ventilation when required was set at pressure support 7 cmH2O and positive end expiratory pressure 0cmH2O (PSV7 +ZEEP). Diaphragm ultrasound was used to assess TF at the end of each step. Demographic, left ventricle ejection fraction (LVEF) assessed through echocardiography and outcome data were collected. Results Thirty-seven patients were screened and 15 patients with VA ECMO were included. Ten patients were extubated, 5 were ventilated in PSV7+ZEEP. The baseline TF was 11% [4 – 15] at a SGF of 2.3 [2 - 4] l/min. TF was 6.3% [0 - 10] at 4 l/min, 13.3% [10 - 26] at 2 l/min and 26.7% [22 - 44] at 1 l/min (ANOVA, P Discussion Diaphragm TF was related with VA ECMO settings and LVEF at the time of the first assessment of patient weaning. Inappropriate ECMO settings lead to impaired diaphragm activity, which could consequently produce atelectasis, pulmonary complications or respiratory muscle fatigue.