Unraveling Predictors of Survival in out-of-hospital-cardiac arrest ECPR: Insights from Principal Component Analysis - The 72nd Annual Conference of the Israel Heart Society in association with the Israel Society of Cardiothoracic Surgery

Yigal Helviz Amir Orlev ² Daniel Fink Frederic S Zimmerman David Shimony Shalva Ben Meir Refael Strugo Reem Naimy Michael Glikson Phillip D Levin

¹Intensive Care Unit, Shaare Zedek Medical Center, Faculty of Medicine, the Hebrew University of Jerusalem, Israel
²Jesselson Integrated Heart Center, The Eisenberg R&D Authority, Shaare Zedek Medical Center, Faculty of Medicine, the Hebrew University of Jerusalem, Israel
³Department of Cardiothoracic Surgery, Shaare Zedek Medical Center, Faculty of Medicine, the Hebrew University of Jerusalem, Israel
⁴Medical division, Magen David Adom, Israeli national EMS organization, Israel

Introduction:
Survival after out-of-hospital cardiac arrest (OHCA) remains low, with less than 10% achieving meaningful neurological recovery. Extracorporeal membrane oxygenation (ECMO) has been introduced as a potential life-saving measure, but its benefit is inconsistent. Key prognostic factors, such as age, initial rhythm, and time to intervention, vary across studies. This study uses principal component analysis (PCA) to assess whether pre-ECMO variables can explain outcomes in OHCA ECPR patients.

Methods:
At Shaare Zedek Medical Center (SZMC), 62 OHCA patients underwent ECPR from 2018–2024. Data collected included age, sex, times from call to EMS/hospital/ECMO, location of arrest, initial pH and lactate, adrenaline doses, and defibrillations. PCA was performed to reduce dimensionality and identify key predictors. Eigenvalues and eigenvectors were used to determine the most influential variables. Survivors were analyzed separately to identify distinct patterns. Permutations and bootstrapping were applied to increase statistical power. Analyses were performed in R 4.3.3.

Results:
Among 62 patients (median age 54, 18% women), the majority experienced myocardial infarction (67%). Median time from call to EMS was 9 minutes, with a median call-to-hospital time of 40 minutes. ECMO weaning was successful in 31%, and 15% survived to hospital discharge.PCA revealed two primary components explaining 51.4% of the variance. PC1 correlated strongly with time to hospital, adrenaline dosage, and shocks, while PC2 correlated with floor of arrest and time to EMS arrival. A reanalysis focused on survivors showed PC1 emphasizing time to EMS arrival inversely related to adrenaline and shocks, and PC2 highlighting distance inversely related to arrest location.

Conclusions:
PCA identified differences in key variables between survivors and non-survivors, though variance explained remains limited. These findings suggest survival depends on a complex interplay of factors. Further research is required to refine predictors and optimize patient selection for ECPR.