A single center experience and outcomes with central and peripheral cannulations for paediatric extracorporeal cardiopulmonary resuscitation.

Abstract

Cannulation metrics for paediatric extracorporeal cardiopulmonary resuscitation (ECPR) have not been widely reported. We hypothesized that duration of resuscitation and cannulation would be associated with survival, and that difficulty and failure rates may differ between peripheral and central approaches. This single-center retrospective study included patients < 18  years with in-hospital ECPR events from 2015 to 2020. The primary and secondary outcomes were survival and favourable neurologic outcomes on hospital discharge. We also set out to evaluate durations of resuscitation and cannulation, characteristics of difficult and failed cannulations, and report the context and mode of death. Statistical methods included non-parametric testing and regression methods. Qualitative assessments were used to describe difficult cannulations. Ninety-two ECPR events met eligibility criteria. Median age was 4 months (IQR 1 month, 16 years), and weight was 4.4 kg (range 1.9-133 kg). Central cannulation was performed in 43% (40/92). Difficult cannulation occurred in 17% (9/52) of peripheral and 13% (5/40) of central cannulations. ECMO flows were achieved in 91% (84/92); failure to achieve flows occurred exclusively with peripheral attempts (N = 8/52). The median (IQR) total resuscitation duration, from time of arrest to ECMO flow initiation (CPA-ECMO), was 31 min (21, 38.6) for central and 36.9 min (31.7, 46.0) for peripheral approaches, P = 0.12. Survival to hospital discharge was associated with the total resuscitation duration (CPA-ECMO) after adjusting for cannulation approach, precipitating event, initial shockable rhythm, and location of cannulation (adjusted OR 0.95, 95% CI 0.92-0.99, p = 0.023). Similarly, the odds of having a favourable neurologic outcome at hospital discharge were associated with the total resuscitation duration (CPA-ECMO) after adjusting for cannulation approach, precipitating event, initial shockable rhythm, and location of cannulation (adjusted OR 0.95, 95% CI 0.92-0.99, p = 0.030). Adjusting for age or patient weight did not change either of these estimates. In paediatric ECPR, total resuscitation duration remains a key metric associated with patient outcomes. In this single-center study, open peripheral cannulation is associated with more difficult and failed attempts compared to central cannulation.