Background: The use of extracorporeal membrane oxygenation (ECMO) for acute respiratory failure after injury is controversial and poorly described.
Methods: We reviewed our single-center experience with use of ECMO from January 2006 to November 2015 at a Level 1 primary adult resource center for trauma to determine the association of in-hospital mortality with patient demographics and clinical variables.
Results: Forty-six patients were treated with ECMO. Patients requiring venoarterial ECMO (n = 7) were excluded. Thirty-nine (85%) were cannulated for venovenous ECMO. Of these, 44% patients survived to discharge. Median age was 28 years. Survivors had a lower BMI and PaCO2 at time of cannulation. Nonsurvivors were more severely injured (median Injury Severity Score, 41 vs. 25; p = 0.03), had a lower arterial pH on arrival, and a shorter length of stay (11 vs. 41 days; p = 0.006). Neither mechanism of injury nor indication for ECMO was associated with mortality. Forty-one percent developed at least one ECMO-related complication, but this was not associated with mortality. Ninety-four percent of the survivors were anticoagulated with heparin versus 55% of nonsurvivors (p = 0.01). Median Injury Severity Score and presence of TBI were not significantly different between survivors and nonsurvivors who were anticoagulated.
Conclusion: The use of venovenous ECMO for acute lung injury after trauma should be considered in special patient populations. Ability to tolerate systemic anticoagulation was associated with improved survival.
Level of evidence: Therapeutic study, level V.