Reports documenting the use of extracorporeal membrane oxygenation (ECMO) after blunt thoracic trauma are scarce. We used a large, multicenter database to examine outcomes when ECMO was used in treating patients with blunt thoracic trauma.METHODS
We performed a retrospective analysis of ECMO patients in the Extracorporeal Life Support Organization database between 1998 and 2014. The diagnostic code for blunt pulmonary contusion (861.21, DRG International Classification of Diseases—9th Rev.) was used to identify patients treated with ECMO after blunt thoracic trauma. Variations of pre-ECMO respiratory support were also evaluated. The primary outcome was survival to discharge; the secondary outcome was hemorrhagic complication associated with ECMO.RESULTS
Eighty-five patients met inclusion criteria. The mean ± SEM age of the cohort was 28.9 ± 1.1 years; 71 (83.5%) were male. The mean ± SEM pre-ECMO PaO2/FIO2 ratio was 59.7 ± 3.5, and the mean ± SEM pre-ECMO length of ventilation was 94.7 ± 13.2 hours. Pre-ECMO support included inhaled nitric oxide (15 patients, 17.6%), high-frequency oscillation (10, 11.8%), and vasopressor agents (57, 67.1%). The mean ± SEM duration of ECMO was 207.4 ± 23.8 hours, and 63 patients (74.1%) were treated with venovenous ECMO. Thirty-two patients (37.6%) underwent invasive procedures before ECMO, and 12 patients (14.1%) underwent invasive procedures while on ECMO. Hemorrhagic complications occurred in 25 cases (29.4%), including 12 patients (14.1%) with surgical site bleeding and 16 (18.8%) with cannula site bleeding (6 patients had both). The rate of survival to discharge was 74.1%. Multivariate analysis showed that shorter duration of ECMO and the use of venovenous ECMO predicted survival.CONCLUSION
Outcomes after the use of ECMO in blunt thoracic trauma can be favorable. Some trauma patients are appropriate candidates for this therapy. Further study may discern which subpopulations of trauma patients will benefit most from ECMO.LEVEL OF EVIDENCE
Therapeutic/care management study, level V.