Abstract 282: Extracorporeal Resuscitation For Out-of-Hospital Cardiac Arrest Patients: Analysis By Time-Dependent Propensity Score Matching Using Nationwide Database In Japan

Abstract

Background: Extracorporeal cardiopulmonary resuscitation (ECPR) has been focused as a rescue therapy for the refractory cardiac arrest patients; however, little is known about the effect of ECPR on survival. The aim of this study is to evaluate the association between ECPR and the survival among the out-of-hospital cardiac arrest patients. Method: This was the secondary analysis of JAAM-OHCA registry which is a nationwide multicenter prospective study collecting the data of out-of-hospital cardiac arrest patients from July 2012 to December 2019. The study participant were adult (≥18 years) cardiac arrest patients due to internal medical cause. Initial cardiac rhythm was stratified as initial shockable and non-shockable rhythm. Patients received ECPR were sequentially matched with the patients who had not yet received within the same minutes (control) based on the time-dependent propensity scores calculated from the patient characteristics to eliminate the resuscitation time bias. Odds ratio with 95% confidence interval (CI) of ECPR for the 30-day survival were calculated by conditional logistic model. Results: Of 57,754 patients in the JAAM-OHCA registry, 1,826 matched cohort with initial shockable rhythm (913 treated with ECPR and 913 control) and 740 matched cohort with initial non-shockable rhythm (370 treated with ECPR and 370 control) were eligible for analysis. Their characteristics between with ECPR and control were well balanced. In the matched cohort with initial shockable rhythm, 30-day survival was 24.6% (225/913) in the patients with ECPR and 16.3% (149/913) in the control. In the matched cohort with initial non-shockable rhythm, 30-day survival was 11.9% (44/370) in the patients with ECPR and 2.4% (9/370) in the control. Odds ratio of with ECPR for 30-day survival was 1.76 [95%CI: 1.38-2.25] in shockable rhythm and 5.37 [95%CI: 2.53- 11.43] in non-shockable rhythm, referred to the control. Conclusion: This time-dependent propensity score sequential matching analysis suggested that ECPR was associated with the 30-day survival outcomes among the out-of-hospital cardiac arrest patients with both initial shockable and non-shockable rhythms.