Refining ECPR candidate selection: from fixed criteria to a dynamic probability-of-benefit framework

Extracorporeal cardiopulmonary resuscitation (ECPR) has emerged as a rescue strategy for selected patients with refractory cardiac arrest. However, identifying patients most likely to achieve neurologically meaningful survival remains challenging. Conventional selection criteria, including age, initial rhythm, witnessed arrest, bystander cardiopulmonary resuscitation (CPR), and no-flow and low-flow duration, are clinically useful but may be insufficient when interpreted as isolated thresholds.

We conducted a narrative review of contemporary evidence on ECPR patient selection, incorporating randomized trials, observational studies, registries, systematic reviews and meta-analyses, and current guideline and consensus recommendations. Evidence was critically evaluated with particular attention to patient-level prognostic factors, arrest setting, dynamic physiological and metabolic variables, ECPR logistics, and system-level determinants of outcome.

Current evidence supports a multidimensional rather than single-variable approach to ECPR candidacy. Favorable arrest characteristics, potentially reversible etiology, shorter low-flow duration, and preserved physiological reserve remain important; however, chronological low-flow time should be interpreted alongside the quality and effectiveness of ongoing CPR. Dynamic information-including transient return of spontaneous circulation, lactate kinetics, arterial pH, end-tidal carbon dioxide, and hemodynamic parameters-may further refine assessment of biological viability. Frailty, comorbidity burden, cerebral perfusion, and extracorporeal membrane oxygenation-related contraindications should also be incorporated into decision-making. Importantly, selection cannot be separated from system performance: Cannulation strategy, transport logistics, institutional experience, and the ability to establish extracorporeal circulation rapidly may substantially modify the likelihood of benefit. Differences between out-of-hospital and in-hospital cardiac arrest further limit the universal application of rigid eligibility criteria.

Candidacy for ECPR should be viewed as a dynamic, probability-based assessment rather than a checklist of fixed thresholds. Integrating arrest characteristics, reversibility, evolving physiology, patient reserve, and system capability may provide a more clinically meaningful framework for identifying patients most likely to achieve neurologically favorable survival after ECPR.