Prevalence and Neurological Outcomes of Comatose Patients withExtracorporeal Membrane Oxygenation

Background : Coma and disorders of consciousness is understudied in patients on extracorporeal membrane oxygenation (ECMO). The objective of our study was to investigate the prevalence, risk factors, and in-hospital outcomes of comatose ECMO patients. Method: This is a retrospectively observational cohort study in a tertiary academic hospital. All adults (age≥18) who received venoarterial (VA) or venovenous (VV) ECMO support between 11/2017 and 04/2022 were included. We defined "24-hour off sedation" as no sedative infusion (except dexmedetomidine) or paralytics administration over a continuous 24-hour period while on ECMO. "Off-sedation coma" (coma off ) was defined as GCS ≤8 after achieving 24-hour off sedation. "On-sedation coma" (coma on ) was defined as GCS≤8 during the entire ECMO course without off-sedation for 24 hours. Neurological outcomes were assessed at discharge using the modified Rankin scale (good 0-3 and poor 4-6). Results: The cohort consisted of 230 ECMO patients (VA-ECMO 143, median age 54, male 65%). "24-hour off sedation" was achieved in 32.2% VA-ECMO and 26.4% VV-ECMO patients. Among all patients off sedation for 24 hours (n=69), 56.5% VA-ECMO and 52.2% VV-ECMO patients experienced coma off . Among those unable to be sedation-free for 24 hours (n=161), 50.5% VA-ECMO and 17.2% VV-ECMO had coma on . Coma off was associated with poor outcomes (p<0.05) in both VA-ECMO and VV-ECMO groups while coma on only impacted the VA-ECMO group outcomes. In a multivariable analysis, the number of packed red blood cell (pRBC) transfusion (aOR=1.16, 95% CI=1.04-1.28), average lactate level (aOR=1.91, 95% CI=1.11-3.30), and acute brain injury (ABI) (aOR=6.41, 95% CI=1.17-35.26) during ECMO support were independent risk factors for coma off after adjusting for renal replacement therapy, ECMO configuration (VA vs. VV), worst pre-ECMO PaO 2 and PaCO 2 . Conclusions : Coma off was common in ECMO patients and was associated with poor neurological outcomes at discharge. The number of pRBC transfusions, high lactate levels, and ABI were independent risk factors.