Background: Veno-venous extracorporeal membrane oxygenation (ECMO) is considered an established rescue therapy in severe respiratory failure. Current knowledge on the optimal ventilation strategy during ECMO is sparse. In recent years, the concept of ultra-protective ventilation has been emerging, aiming to reduce ventilator-induced lung injury. This study sought to investigate ventilator settings used in patients on ECMO and their impact on mortality. Methods: Retrospective Analysis of patients undergoing invasive mechanical ventilation and simultaneous ECMO between 01/2009 and 06/2016. Patients on non-invasive ventilation, veno-arterial ECMO, or low-flow ECMO for carbon dioxide removal were excluded. Results: 62 patients were included (age 49.45 ±17.79 a, 59.7% male). ECMO was administered using a blood flow of 3.55 ± 0.93 lpm and a sweep-gas flow of 3.37 ±1.51 lpm. Upon ECMO initiation, tidal volumes could be reduced from 6.25 ±2.63 to 4.13 ±2.10 ml/kg PBW. Changes were due to reduction of the inspiratory plateau pressure (29.2 ±5.85 vs. 22.7 ±3.84 cmH2O), thus reducing driving pressure (19.25 ±6.35 vs. 12.25 ±3.48 cmH2O). Positive end-expiratory pressure remained constant (9.82 ±3.81 vs. 10.59 ±3.70 cmH2O). Despite ultra-protective ventilation, gas exchange during ECMO was sufficient (pO2 79.60 ±19.60 mmHg, pCO2 46.66 ±9.19 mmHg). 33 patients (53.2%) died. There were no significant differences in ventilator parameters between survivors and non-survivors. Conclusion: Our data show that ultra-protective ventilation during ECMO therapy can be feasibly applied under real-world conditions. However, a positive prognostic effect could not be detected.
