Seizures in children undergoing extracorporeal membrane oxygenation: a systematic review and meta-analysis

Objectives To investigate the incidence of seizures and short-term mortality associated with seizures in children undergoing extracorporeal membrane oxygenation (ECMO). Methods PubMed, Embase, and Web of Science were searched from inception to September 2021. Study quality was assessed using the Newcastle-Ottawa Scale. Random effects meta-analysis was conducted. Results Fourteen studies met the inclusion criteria for quantitative meta-analysis. The cumulative estimate of seizure incidence was 15% (95% CI: 12–17%). Studies using electroencephalography reported a higher incidence of seizures compared with those using electro-clinical criteria (19% vs. 9%, P = 0.034). Furthermore, 75% of seizures were subclinical. Children receiving extracorporeal cardiopulmonary resuscitation (ECPR) exhibited a higher incidence of seizures compared to children with respiratory and cardiac indications. Seizure incidence was higher in patients undergoing venoarterial (VA) ECMO compared with venovenous (VV) ECMO. The pooled odds ratio of mortality was 2.58 (95% CI: 2.25–2.95) in those developed seizures. Conclusion The incidence of seizures in children requiring ECMO was 15% and majority of seizures were subclinical. The incidence of seizures was higher in patients receiving ECPR than in those with respiratory and cardiac indications. Seizures were more frequent in patients undergoing VA ECMO than VV ECMO. Seizures were associated with increased short-term mortality. Impact The incidence of seizures in children undergoing extracorporeal membrane oxygenation (ECMO) was ~15% and majority of the seizures were subclinical. Seizures were associated with increased short-term mortality. Risk factors for seizures were extracorporeal cardiopulmonary resuscitation and venoarterial ECMO. Electroencephalography (EEG) monitoring is recommended in children undergoing ECMO and further studies on the optimal protocol for EEG monitoring are necessary.