Prehospital Administration of Tranexamic Acid in Trauma Patients: A 1:1 Matched Comparative Study From a Level 1 Trauma Center
Authors: El-Menyar A, Sathian B, Wahlen BM, et al.
Published in: Am J Emerg Med, 2019 Apr 30.
Over the past few years there has been a fair amount of discussion regarding the use of tranexamic acid (TXA) in the prehospital environment. TXA is said to benefit trauma patients with hemodynamic instability and noncompressible bleeding. It’s a relatively affordable intervention, and studies have shown it can reduce mortality. Although the European guideline on management of major bleeding and coagulopathy trauma recommends early administration of TXA to prevent bleeding, the empirical evidence supporting its administration in the prehospital environment is limited.
The authors of this month’s study had an objective to evaluate “the efficacy and impact of TXA administration in the prehospital setting on reducing mortality, associated thromboembolic events, and blood transfusion in trauma patients.”
I’m sure you noticed I mentioned European guidelines above. In this column we typically review research completed in the U.S. However, this month’s study was completed in Qatar.
It’s important to review prehospital research from all over the world, particularly in EMS systems comparable to ours. The authors note that in Qatar EMS teams use evidence-based treatment protocols and ambulances are staffed by critical care paramedics and emergency medical technicians. Further, the data for this study was obtained from the Qatar National Trauma Registry, which participates in the American College of Surgeons Committee on Trauma’s National Trauma Data Bank and Trauma Quality Improvement Program.
This was a retrospective study that included all trauma patients from 16 to 80 years old who received 1 g of intravenous TXA prehospitally and were transported to a level 1 trauma center. The study period was from January 1, 2017 to September 30, 2018.
TXA was administered to adult patients with ongoing “significant hemorrhage,” which was defined as systolic blood pressure less than 90 mmHg and/or a heart rate of more than 110 bpm, or to patients at risk of significant hemorrhage. The trauma must also have occurred within three hours of the EMS response. Patients were excluded if they received their first dose of TXA in the ED, were younger than 16 or older than 80, were prisoners, were pregnant, had a traumatic brain injury with exposed brain, or were isolated drowning or hanging victims.
The study outcomes of interest were mortality, “massive transfusion” (defined as 10 or more units of red blood cells over a 24-hour period or more than 40 ml/kg in two hours or less), activation of the massive transfusion protocol (coolers containing six units of uncross-matched type O positive pure red blood cells; six units of equivalent of platelets; and six units of AB plasma were sequentially sent to the resuscitation area), and thromboembolic events including deep vein thrombosis.
There are a few ways investigators attempt to control for confounding variables when performing an analysis like this. Confounding variables are unmeasured variables that could potentially have an impact on the relationship being evaluated. For example, if you are evaluating the relationship between activity level and gaining weight, you would also want to account for age and sex. However, if you’re unable to collect information on age or sex, you can’t adjust for them in the analysis. There could be unknown variables that also impact this relationship. It is important to try to account for confounding variables.
To do this the authors “matched” patients by age, sex, injury severity score, head abbreviated injury score, prehospital heart rate, and systolic blood pressure. The idea behind matching is that the groups being compared are similar in the matched characteristics and therefore should be similar, in theory, with respect to the distribution of unmeasured (confounding) variables. The authors indicated their matching methodology yielded two similar cohorts without statistically significant differences in age, sex, prehospital SBP, DBP, pulse rate, GCS, and shock index (heart rate/SBP).
The authors intended to analyze every patient who met inclusion criteria and had TXA administered during the study period. Therefore, they did not calculate a sample size prior to beginning their study. This has some implications we will discuss shortly.
There were 204 patients identified after matching (102 in the TXA group and 102 in the control group). The study found TXA administration in the prehospital environment was associated with significantly lower rates of massive transfusion protocol activation when compared to the control group (26.5% vs. 56.9%; p=0.001) and significantly lower rates of massive blood transfusion (18.6% vs. 39.2%; p=0.001). There was no statistically significant relationship noted with respect to mortality or thromboembolic events (p>0.05).
While the results weren’t statistically significant, mortality was lower in the TXA group. It was mentioned earlier that the authors did not perform a sample size calculation. This had big implications for the mortality results they reported: A post-hoc analysis (performed after data was collected and analyzed) revealed the study was underpowered to detect a statistical difference in mortality if one was present. In other words, they didn’t have enough data to properly evaluate mortality. Other limitations noted by the authors included the inability to collect data on hyperfibrinolysis, inflammation, organ failure, and sepsis. The authors also noted that those who received TXA also had shorter EMS total call times (62 vs. 74 minutes).
This was an interesting study that adds to the literature on prehospital TXA administration. Its lessons for researchers include making sure your study is able to appropriately evaluate your outcomes of interest. It also emphasizes the need for more research into prehospital TXA administration, which should include clinical trials.
Antonio R. Fernandez, PhD, NRP, FAHA, is research director at the EMS Performance Improvement Center and an assistant professor in the Department of Emergency Medicine at the University of North Carolina–Chapel Hill. He is on the board of advisors of the Prehospital Care Research Forum at UCLA.