The incidence of venous thromboembolic events in trauma patients after tranexamic acid administration: an EAST multicenter study
Lisbi Rivas, Jordan M. Estroff, Andrew Sparks, Jeffry Nahmias, Rebecca Allen, Savannah R. Smith, Matthew Kutcher, Kristen Carter, Areg Grigorian, Spencer Albertson, David Turay, Juan C. Quispe, Xian Luo‐Owen, Michael A. Vella, José L. Pascual, Gabriella Tororello, McKell Quattrone, Andrew C. Bernard, Asanthi Ratnasekera, Alice A. Lee, Danielle Tamburrini, Carlos J. Rodríguez, Kelly M. Harrell, Kokila Jeyamurugan, Nikolay Bugaev, Anne Warner, Jason Weinberger, Joshua P. Hazelton, Mariam Selevany, Franklin L. Wright, Alexandra Kovar, Shane Urban, Amy V. Hamrick, Michael Mount, Matthew Carrick, Daniel C. Cullinane, Grace Chang, Gary Jain, Chance Spalding, Babak Sarani
Abstract
To determine if there is a significant association between administration of tranexamic acid (TXA) in severely bleeding, injured patients, and venous thromboembolism (VTE), myocardial infarction (MI), or cerebrovascular accident (CVA). A multicenter, retrospective study was performed. Inclusion criteria were: age 18-80 years old and need for 5 units or more of blood in the first 24 h after injury. Exclusion criteria included: death within 24 h, pregnancy, administration of TXA more than 3 h following injury, and routine ultrasound surveillance for deep venous thrombosis. Incidence of VTE was the primary outcome. Secondary outcomes included MI, CVA, and death. A power analysis found that a total of 830 patients were needed to detect a true difference in VTE risk. 1333 patients (TXA = 887, No-TXA = 446 patients) from 17 centers were enrolled. There were no differences in age, shock index, Glasgow coma score, pelvis/extremity abbreviated injury score, or paralysis. Injury severity score was higher in the No-TXA group. Incidence of VTE, MI, or CVA was similar between the groups. The TXA group required significantly less transfusion (P < 0.001 for all products) and had a lower mortality [adjusted odds ratio 0.67 (95% confidence interval 0.45-0.98)]. Despite having a higher extremity/pelvis abbreviated injury score, results did not change when evaluating only patients with blunt injury. Use of TXA in bleeding, injured patients is not associated with VTE, MI, or CVA but is associated with a lower transfusion need and mortality.