Search

Background. Extracorporeal membrane oxygenation (ECMO) serves as a rescue therapy when systemic hypoxia persists despite conventional care for severe acute respiratory distress syndrome (ARDS). Due to the extracorporeal gas exchange, the paO2/FiO2 ratio cannot be used as the primary marker for disease severity and progression. Therefore, we performed a propensity score-matched analysis to identify other potential predictors of outcomes in patients supported by ECMO therapy. Results. Between December 2014 and May 2018, 105 patients underwent venovenous ECMO in our institution. From these patients, we identified 28 who died during ECMO therapy and assigned 28 control patients using propensity score matching based on the following criteria: age, ARDS severity, and SAPSII score at admission. A statistical evaluation of the patient characteristics, intensive care data, morbidities, respiratory system variables, and outcomes was performed. The baseline patient characteristics did not differ between groups and ECMO was placed on day 1 in all patients. The analyzed variables of respiratory mechanics, such as the plateau pressure, positive end-expiratory pressure, and tidal volume, did not differ between groups. The driving pressure before ECMO was equal between the nonsurvivors and the controls. Twelve hours after initiation of ECMO therapy, the driving pressure decreased by 40.8% in the survivors but by only 20.1% in the nonsurvivors. Conclusions. We report that very early driving pressure changes can serve as an indicator of disease severity and predict patient survival following ECMO therapy.