54.04 Predicting Survivability of Non-Cardiac Pediatric Patients Requiring eCPR

S. Dixon1, S. Koenig1, M. McBride1, R. Russell1, S. Anderson1, C. Onwubiko1, E. Sparks1  1University Of Alabama at Birmingham, Pediatric Surgery, Birmingham, Alabama, USA

Introduction: Application of Extracorporeal Life Support (ECLS) during cardiac arrest is termed Extracorporeal Cardiopulmonary Resuscitation (eCPR). Provision of eCPR in pediatric patients with primary pulmonary disease is increasing. Mortality in this population remains high and factors influencing survival are not well defined. Our single center study examines the correlation between pre-ECMO lab values and survival to discharge in non-cardiac eCPR patients.

Methods:  A retrospective review of ECMO data at our free-standing children’s hospital from January 2013 through December 2023 was performed. The variables evaluated included age, diagnosis, weight, blood gas values, type of cannulation, CPR time, and survival to discharge. Values outside of the normal range (e.g. pH < 6.77) were approximated to the nearest whole value (e.g. pH = 6.77).  CPR time included resuscitative efforts before and during ECMO cannulation. CPR ceased once the patient was successfully placed on ECMO or patient expired. Kruskal-Wallis Test was used to compare pre-cannulation labs and CPR minutes versus survival to discharge.

Results: There were 23 patients identified over a 10-year period that underwent CPR during ECMO cannulation (14 neonatal & 9 pediatric patients). The median age of patients was 8 days (IQR 2-438 days). The median weight of the patients was 3.77 kg (IQR 2.9-11.45 kg). The most common diagnosis among these patients was persistent pulmonary hypertension (n = 5, 21.7%). The median duration of CPR prior to successful ECMO flows was 60 minutes (IQR 15-80 min). Most of these patients were cannulated onto venous-arterial (VA) ECMO (n = 20, 87%). The median ECMO run time was 84 hours (IQR 27 – 183 hours). A single patient died during ECMO cannulation (n = 1, 4.3%). Eight patients survived to discharge (34.8%). All patients who survived to decannulation also survived to discharge. Most blood gases obtained were arterial (n = 16, 70%), though venous and capillary were also used. Of the eight patients who survived only one (n = 1, 12.5%) of the patients had a recorded pH less than 7.0. Higher pH, higher PaO2, and higher bicarbonate levels prior to cannulation were associated with survival (p < 0.05, see table below). Other characteristics such as lactate and PaCO2, as well as duration of CPR, were not associated with survival.

Conclusion: A higher pH, PaO2, and bicarbonate prior to eCPR were significantly associated with survivability in our pediatric eCPR cohort. Pre-cannulation characteristics that may influence survivability have the potential to assist with decision making regarding inclusion and exclusion criteria for eCPR candidates.