Relationship between 30 Days Mortality and Incidence of Intraoperative Cardiac Arrest According to the Timing of ECMO

Background: Recently, the use of extracorporeal membrane oxygenation (ECMO) in noncardiac surgery, such as thoracic surgery, has increased. However, there have been no studies on the mortality and incidence of intraoperative cardiac arrest with or without ECMO during thoracic surgery. Methods: Between January 2011 and October 2018, 63 patients received ECMO support during thoracic surgery. All patients who applied ECMO from starting at any time before surgery to the day of surgery were included. Patients were divided into the emergency ECMO group and the non-emergency ECMO group according to the timing of ECMO. We compared the factors related to 30 day mortality using Cox regression analysis. Results: The emergency ECMO and non-emergency ECMO groups comprised 27 and 36 patients, respectively. On the operation day, cardiopulmonary resuscitation (CPR) was a very important result, and only occurred in the emergency ECMO group (n = 20, 74.1% vs. 0%, p < 0.001). The most common cause of ECMO indication was the CPR in the emergency ECMO group and respiratory failure in the non-emergency ECMO group. There were significant differences in 30 day mortality between the emergency ECMO group and the non-emergency ECMO group (n = 12, 44.4% vs. n = 3, 8.3%, p = 0.001). The Kaplan–Meier analysis curve for 30 day mortality showed that the emergency ECMO group had a significantly higher rate of 30 day mortality than the non-emergency ECMO group (X(2) = 14.7, p < 0.001). Conclusions: A lower incidence of intraoperative cardiac arrest occurred in the non-emergency ECMO group than in the emergency ECMO group. Moreover, 30 day mortality was associated with emergency ECMO.