Optimizing patient selection for ECMO after pediatric hypothermic cardiac arrest

In 2004, our level 1 regional pediatric trauma center created a protocol to activate ECMO for children with suspected hypothermic cardiac arrest based on inclusion criteria: serum potassium ≤9, submersion <90 min, and core body temperature <30 °C. In 2017, Pasquier et al. developed a model to help predict the survival of adults after hypothermic cardiac arrest (HOPE score) that has not been validated in children. We sought to apply this score to our pediatric patient population to determine if it can optimize our patient selection.

This was a retrospective review of all patients cannulated onto VA ECMO for hypothermic cardiac arrest between 2004 and 2022. We used abstracted data points to calculate the HOPE score for our patient population, both with and without presumed asphyxia.

Over 19 years, 18 patients were cannulated for suspected hypothermic arrest, with three survivors (17 %). The HOPE score survival prediction ranged from 1 to 86 % with presumed asphyxia and 6-98 % without presumed asphyxia. Survivor HOPE scores ranged from 9 to 86 % with presumed asphyxia and 42-98 % without presumed asphyxia. Non-survivors' scores ranged 1-29 % with asphyxia and 6-57 % without asphyxia. A cutoff of >5 % predicted survival with asphyxia for ECMO could have decreased our cannulations by half without missing survivors.

ECMO can be a lifesaving measure for specific children after hypothermic arrest. However, identifying the patients that will benefit from this resource-intensive intervention remains difficult. HOPE score utilization may decrease the rate of futile cannulation in children, but multi-centered research is needed in the pediatric population.