Methods and Materials:
A retrospective cohort review was performed on all entries in the Extracorporeal Life Support Organization (ELSO) registry from November 1, 2010 to July 31, 2017 at the Massachusetts General Hospital (MGH), a large quaternary care center in Boston, Massachusetts (USA). Patients who were undergoing CPR in hospital at the time of ECMO initiation as documented in the procedural report were included. Patients who underwent ECMO cannulation but for whom immediate durable flow through the ECMO circuit could not be initiated were excluded. Baseline characteristics were extracted from the medical record. The study was approved by the Partners Institutional Review Board.
This study’s primary outcome was survival to hospital discharge. Etiology of cardiac arrest was adjudicated by two reviewers (NM & LXM) and categorized by cardiac origin ( defined as an etiology of acute coronary syndrome, primary arrhythmia in the absence of overlapping etiology, acute heart failure, and tamponade) or non-cardiac origin (defined as pulmonary embolism, drug toxicity, sepsis, perioperative shock, or air embolism). When multiple potential etiologies were present, the primary cause as adjudicated by the care team was selected. Cardiac arrest rhythm was ascertained from the procedural report and categorized as either pulseless ventricular tachycardia/ventricular fibrillation (VT/VF) or pulseless electrical activity (PEA).
ECMO-related complications were assessed. Major bleeding was defined as Bleeding Academic Research Consortium category 3-527. Additional outcomes assessed included de novo renal replacement therapy, cerebrovascular events, limb ischemia, and cerebral performance category (CPC). CPC scores 1 through 5 correlated to good cerebral performance (able to work, minor deficits), moderate cerebral disability (persistent deficits but mostly independent), severe cerebral disability (dependent on others), coma or vegetative state, and brain death, respectively28.
Descriptive statistics were used to quantify baseline patient characteristics, clinical features, ECMO utilization, and survival rates. Categorical variables were reported as counts and percentages and compared using the Fisher’s exact test. Continuous variables were reported as medians with interquartile ranges and compared with the Wilcoxon two-sample test. Statistical significance was assessed at a nominal α level of 0.05. All reported P values were 2-sided. Analysis was performed in R 4.0.0 (R Core Team, 2020).