Prone Positioning Under ECMO: A Retrospective Cohort of 34 Patients

Abstract

<h3>Purpose</h3> Extracorporeal Membrane Oxygenation (ECMO) become standard therapy for refractory hypoxemia (veno-venous (VV)) and circulatory failure (veno-arterial (VA)). However, severe hypoxemia may persist in some patients under ECMO. Prone position (PP) has been shown to have a benefit on oxygenation in patients with acute respiratory distress syndrome. We performed a retrospective study to evaluate the change of oxygenation and the risks involved in positioning patients under ECMO in PP. <h3>Methods</h3> All patients who underwent PP under ECMO from August 2014 to December 2020 were included. PP was performed on two criteria: refractory hypoxemia defined by a P/F<80mmHg despite 100% FiO2 on ventilator and ECMO or persistent hypoxemia defined by a need to maintained FiO2 ≥80% on ECMO with posterior pulmonary condensations on CT scan. Ventilatory and ECMO parameters and PP-related complications were collected. <h3>Results</h3> From 556 patients under ECMO, 34 (6.1%) (median age 54 years) were included for 87 PP procedures (2.5 +/-2.4 sessions per patient). ECMO was VA for 6 patients, VV for 27 patients and VAV for one patient. 19 patients were placed in PP for refractory hypoxemia and 15 for persistent hypoxemia. Mean tidal volume was 4.5mL/kg +/-1.7, median positive end-expiratory pressure at 10 cmH20 (IQR 10-15), respiratory rate at 20/min (IQR 18-28). Mean PP session duration was 18 h +/-4.2. Median time from ECMO placement to first PP was 7 days (IQR 4-10). PP procedures resulted in a significant improvement in oxygenation, with increasing P/F_ECMO ratio from 92.6 to 133.2 on average (p<0.0001). Concomitantly the ventilator FiO2 decreased from 65% to 55% (p<0.0001).The most frequent complication was the presence of bedsores of the face and trunk (6 patients). The other complications reported were: 1 cardio respiratory arrest due to tamponade, 1 pump thrombosis, 1 displacement of limb reperfusion cannula, 1 tracheostomy decanulation. No ECMO decanulation occurred. At discharge, 9 of 34 patients with PP (26%) had died compared to 237 on 522 patients under ECMO without PP(45.4%) (p=0.033). <h3>Conclusion</h3> PP in hypoxemic patients under ECMO is effective for improving oxygenation. A beneficial effect on mortality is also suggested. In an expert center, less than 20% of patients experienced bedsores and 12% may suffer from severe complications requiring urgent treatment.