ECMO procedure
Two operators performed ECMO insertion to patients during CPR in our institute. Arterial (15 and 17Fr, Bio-medicus, Medtronic Inc. MN, USA) and venous (20, 22, and 24Fr Edwards Lifescience Inc., Irvine, CA, USA) cannula were inserted through a femoral artery and vein by the percutaneous Seldinger technique without ultrasonographic or fluoroscopic guidance. Heparin was administered immediately before cannulation at 50 IU/kg and continuously infused after ECPR, and activated prothrombin time was maintained between 1.3-1.8 times normal range. Two types of ECMO consoles were used (Capiox Emergency bypass system; Terumo Inc., Tokyo, Japan, and Permanent Life Support; Maquet Cardiopulmonary AG, Rastatt, Germany). Left heart decompression was achieved using a left atrial catheter (22, and 24Fr Edwards Lifescience Inc., Irvine, CA, USA) via percutaneous atrial septostomy through a femoral vein. When leg ischemia was confirmed, distal perfusion was achieved by inserting a 16 gauge central venous catheter (ARROW/Teleflex, Wayne, USA) into a superficial femoral artery; the catheter was then attached to the side port of the return cannula. “Awake ECMO” was attempted without mechanical ventilation support when a patient could spontaneously maintain breathing. In some patients, where the mean arterial blood pressure (MAP) or pulse pressure could not be maintained during ECMO support, inotropes such as dobutamine, and isoproterenol were administered. MAP was controlled according to patient perfusion status after checking lactate level – usually, target MAP ranged from 60-80 mmHg. When MAP could not be maintained in the target range despite full ECMO support, inotrope infusion was initiated. At this time, either isoproterenol or dobutamine was chosen if pulse pressure had disappeared along with decline in MAP. On the other hand, if pulse pressure was maintained, vasopressors such as norepinephrine or vasopressin were initiated. Weaning off ECMO was commenced when cardiac function improvement was confirmed by transthoracic echocardiography. A reduction in ECMO flow by 50 % was attempted for more than 30 minutes if hemodynamic stability was maintained. Optimized inotropic support was considered for patients exhibiting a fall in MAP of more than 10 mmHg during the 50% ECMO support period. Because the blood in the ECMO circuit was discarded, two units of packed red blood cells were transfused into all patients when ECMO was weaned.