Cardiac rhythm is a well-established prognostic factor in patients undergoing extracorporeal cardiopulmonary resuscitation (ECPR); however, existing evidence is primarily limited to small, single-center studies. This study aimed to evaluate the prognostic value of both initial rhythm and rhythm at the time of cannulation using data from the international Extracorporeal Life Support Organization (ELSO) Registry.

We analyzed adult ECPR cases (≥18 years) recorded in the ELSO Registry between 2020 and 2025. The primary outcome was survival to hospital discharge. Multivariable logistic regression was used to assess associations between survival and cardiac rhythm (initial and at cannulation), adjusting for age, sex, witnessed arrest, arrest location, and CPR duration.

Among 5489 adult ECPR cases, 3382 were included in the complete case analysis. Median age was 57.1 years (IQR 45.7–65.5), 29.1 % were female, and median CPR duration was 44 min (IQR 28–61). Survival proportions were highest in patients with persistent ventricular fibrillation (VF) or conversion from an initial non-shockable rhythm to VF, and lowest in those who either converted to or remained in non-shockable rhythm. In multivariable analysis, initial pulseless electrical activity (PEA) (OR 0.59, 95 % CI 0.48–0.71) and asystole (OR 0.64, 95 % CI 0.47–0.86) were associated with lower odds of survival compared to initial VF. Asystole at the time of cannulation was associated with worse outcomes (OR 0.61, 95 % CI 0.38–0.99). The survival decline with prolonged CPR was steepest in those with asystole at cannulation (OR for 60 vs 40 min of CPR = 0.64; CI 0.54–0.76; p < 0.001) and most gradual in patients with VF at cannulation (OR for 60 vs 40 min of CPR = 0.84; CI 0.76–0.93; p = 0.001).

In this large, multicenter analysis, both initial and cannulation rhythms were independently associated with survival among ECPR recipients. Survival showed time-dependent variation across rhythm conversions. These specific survival patterns underscore the need for individualized, rhythm-informed ECPR decision-making algorithms to optimize patient outcomes.