Association of Tranexamic Acid Administration With Mortality and Thromboembolic Events in Patients With Traumatic Injury

Key Points Question Is the administration of tranexamic acid associated with mortality and thromboembolic events in patients with traumatic injury? Findings This systematic review and meta-analysis included 31 randomized clinical trials and observational studies with 43 473 patients, among whom tranexamic acid was associated with lower 1-month mortality compared with the control cohort. Data from tranexamic acid studies, especially thromboembolic events, are largely heterogeneous. Meaning These findings suggest that tranexamic acid may decrease mortality in patients with traumatic injury, but the incidence of thromboembolic events remains unclear because trials and populations with traumatic injury are often not comparable and data vary largely.


Selection process
Titles and abstracts of records identified by the electronic literature search were screened first. Next, full text articles were retrieved and screened for eligibility criteria. Disagreements were solved through discussion between the two reviewers or after consultation with a third or fourth reviewer until consensus. Multiple reports of the same study were merged so that each study is the unit of analyses.

Data items extracted
© 2022 Karl V et al. JAMA Network Open.
We extracted data of the latest available follow-up time for the outcomes overall mortality and thromboembolic events. As there are some studies which conducted a subgroup analysis within their trials, there are cases in which we extracted different data from the same study according to the subgroup. Missing data were marked as such and evaluated in the bias assessment and discussion, respectively. Extracted data items were extracted as follows.

Study selection
Our systematic literature search identified 1546 records. Two additional studies were detected through manual reference screening of related articles. After removal of duplicates, 1317 abstracts were screened for inclusion and exclusion criteria. 71 publications were included for full text screening, at which stage 40 studies were excluded (eTable 1). We included 31 studies in the systematic review.

Outcomes
During study selection it became apparent that some studies did not report the time at which mortality was measured (follow-up time), whereas others reported follow-up times that differed from the above. In order to avoid excluding many studies due to incorrect follow-up times, we suggested overall mortality in addition. This outcome encompassed mortality of all aforementioned time points plus mortaltiy for which no time frame was given.
It was not possible to conduct a meta-analysis for the amount of blood products given. A lot of data on blood product transfusion were reported as median instead of mean. Moreover, due to major clinical differences between the studies regarding trauma severity, data varied largely and led to significant heterogeneity. Thus, we had to discard the outcome amount of blood products given. Detailed data of blood products given can be seen in the Supplement XIII.

Results of subgroup analyses
After extracting data it was impossible to conduct meta analyses of some subgroups. First, most of the studies included patients that were severely injured, therefore we did not have sufficient data for the non-severely injured cohort to compare. Second, we are missing studies that compare blunt vs. penetrating trauma and TXA administration within 3 hours vs. beyond 3 hours, respectively. Finally, we conducted meta-analyses of the subgroups multiple trauma vs. TBI and in-hospital vs. prehospital TXA administration. For the latter, data availability limited our analysis to the outcomes overall mortality and thromboembolic events.
After recognising that most of our results were statistically heterogenous, we hypothesized that different TXA dosage regimens between trials would be a reason. Some included studies did not define a dosage regimen prior to study initiation while others applied a dosage regimen that differs from the The European guideline on management of major bleeding and coagulopathy following trauma: 1g IV bolus, followed by 1g IV maintenance dose over 8 hours 38 . Therefore we conducted post-hoc meta-analyses comparing standard TXA administration and non-standard administration.    Intervention 1 g TXA on admission followed by 500 mg every 6 hours for 24 hours

Control
No TXA

Time of TXA administration after injury
On admission • First dose of TXA at the ED • < 16 and > 80 years of age • Vulnerable population (prisoners and pregnant patients) • Traumatic brain injury with exposed brain • Isolated drowning or hanging victims

Inclusion criteria
• Blunt or penetrating trauma with signs and symptoms of hemorrhagic shock within three hours of injury (SBP less than 90 mmHg at scene of injury, during air and/or ground medical transport, or upon arrival to designated trauma centers; Heart rate > 120; Estimated blood loss of 500 milliliters in the field.

Control
No TXA • Bleeding not controlled by direct pressure or tourniquet • Major amputation of any extremity above the wrists and above the ankles

Exclusion criteria
• Any patient <18 years of age • Any patient more than three hours post-injury • Any patient with an active thromboembolic event (within the last 24 hours)i.e., active stroke, myocardial infarction or pulmonary embolism • Any patient with a hypersensitivity or anaphylactic reaction to TXA • Traumatic arrest with more than five minutes of cardiopulmonary resuscitation without return of vital signs • Penetrating cranial injury • Traumatic brain injury with brain matter exposed • Isolated drowning or hanging victims • Documented cervical cord injury with motor deficits

Inclusion criteria
• Blunt or penetrating trauma with signs and symptoms of hemorrhagic shock within three hours of injury (SBP of less than 90 mmHg upon arrival to designated trauma centers, HR > 120, estimated blood loss of 500 milliliters, bleeding not controlled by direct pressure or tourniquet • Major amputation of any extremity above the wrists and above the ankles

Exclusion criteria
• Any patient <18 years of age • Any patient more than three hours post-injury • Any patient with an active thromboembolic event (within the last 24 hours)i.e., active stroke, myocardial infarction or pulmonary embolism • Any patient with a hypersensitivity or anaphylactic reaction to TXA • Any patient that received prehospital TXA • Traumatic arrest with more than five minutes of cardiopulmonary resuscitation without return of vital signs • Penetrating cranial injury • Traumatic brain injury with brain matter exposed • Isolated drowning or hanging victims • Documented cervical cord injury with motor deficits