Tranexamic Acid in Civilian Trauma Care in the California Prehospital Antifibrinolytic Therapy Study

Michael M. Neeki, DO, MS; Fanglong Dong, PhD; Jake Toy, BA; Reza Vaezazizi, MD; Joe Powell, EMT-P; David Wong, MD; Michael Mousselli, BS; Massoud Rabiei, BS; Alex Jabourian, DO; Nichole Niknafs, DO; Michelle Burgett-Moreno, BA; Richard Vara, RN, BSN; Shanna Kissel, RN, MSN; Xian Luo-Owen, MD, PhD; Karen R. O'Bosky, MD; Daniel Ludi, MD; Karl Sporer, MD; Troy Pennington, DO; Tommy Lee, MD; Rodney Borger, MD; Eugene Kwong, MD


Western J Emerg Med. 2018;19(6):977-986. 

In This Article


CAL-PAT Study Overview

The Cal-PAT study was a multi-centered, prospective, observational cohort study with a retrospective comparison. The study was initiated in March 2015 in two Southern California counties–San Bernardino and Riverside. In early 2016 Alameda County joined the study. All eight receiving centers are designated Level I and Level II trauma centers. A total of 30 emergency medical services (EMS) agencies were involved across all counties. Current data collection for this study concluded in July 2017 in all counties. Within the prehospital setting, the California Emergency Medical Services Authority approved TXA to be included in EMS protocols as a standard treatment for all trauma patients showing signs of hemorrhagic shock. TXA administration was carried out uniformly among all participating EMS agencies. The institutional review board at each trauma center approved CAL-PAT study protocols, including the incorporation of TXA into the massive transfusion protocol at each center as a standard of care for trauma patients and allowed for research data collection with a waiver of consent.

Data collection, Protocols, Outcomes

All patients ≥18 years old who sustained blunt or penetrating trauma with signs of hemorrhagic shock were considered for TXA treatment upon meeting enrollment criteria (Figure 1). Patient selection in the prehospital setting was determined by paramedics on ambulances or by registered nurses on helicopter transport units. Paramedics and registered nurses underwent a standardized training session including education on the guidelines for TXA candidate identification, the protocol for TXA administration, and the TXA known side-effect profile. Additionally, a system of access to real-time consultation with senior physicians familiar with study protocol at each participating trauma center was established prior to study initiation to address any first responder concerns regarding patient selection or TXA administration.

Figure 1.

Inclusion and exclusion criteria provided to first responders in the field and clinicians at receiving trauma centers. TXA, tranexamic acid.

TXA was delivered in two doses following the protocol used in the CRASH-2 trial.[13,14] The first dose was 1 gram of TXA in 100 ml of 0.9% normal saline infused over 10 minutes via intravenous (IV) or intraosseous access. This first dose was administered by paramedics or registered nurses as soon as feasible after patient assessment. Identification of study patients receiving TXA was achieved through a wristband labeled "TXA", verbal communication at patient hand off by EMS, and/or by EMS run sheet. Following arrival to a participating trauma center, patients who received prehospital TXA were identified and re-assessed by trauma team members for signs of continued hemorrhagic shock. Patients who continued to meet the study criteria (Figure 1) received a second dose of 1 gram of TXA in 100 ml of 0.9% normal saline infused over eight hours via IV infusion. A patient may have received only one dose of TXA if they arrived to the trauma center and no longer met study criteria (Figure 1). We excluded from the study patients who were deceased upon arrival (declared dead on arrival with minimal resuscitation effort or failed to respond to resuscitation after 15 minutes in the ED), those who received TXA for non-trauma indications, and those who received TXA and were determined to be less than 18 years old upon arrival.

The control group was formed of patients seen at each receiving center within five years prior to the conclusion of data collection (June 2012 to July 2017). This group included patients who were not administered TXA because they were brought in by an EMS provider group not carrying TXA or because they were transported to the hospital by any means other than a designated EMS provider (e.g., friends, family, self). The control group patients met the same study criteria (Figure 1) as those in the TXA group. The control group patients were matched to TXA group patients through propensity scoring based upon gender, age, Injury Severity Score (ISS), and mechanism of injury. We further aimed to match TXA group patients with controls from the same trauma center.

The primary outcome was mortality measured at 24 hours, 48 hours, and 28 days. Additional variables included total blood products transfused during the hospital stay, the hospital and intensive care unit (ICU) length of stay (LOS), systolic blood pressure taken prior to TXA administration, Glasgow Coma Score observed prior to the first TXA dose in the field, and the incidence of known adverse events associated with TXA administration including thromboembolic events (e.g., deep vein thrombosis, pulmonary embolism), myocardial infarction, and neurological events (e.g., stroke, seizure).

Data for included subjects were abstracted from the electronic medical record and trauma registry for each patient. Follow up to determine mortality outcomes after hospital discharge was abstracted from the electronic medical record and trauma registry. In select cases, direct chart review was conducted, and in cases of missing data, study investigators contacted patients' and/or patients' families directly to determine survival outcomes. Estimated time to TXA administration by EMS was determined to be the estimated time of injury based on the time that the 911 call was received and documented time of TXA administration on the EMS run sheet.

Statistical Analysis

We conducted all statistical analyses using the SAS software for Windows version 9.3 (SAS Institute, Cary, North Carolina, USA). Descriptive statistics were presented as means and standard deviations for continuous variables, along with frequencies and proportions for categorical variables. Propensity score matching based on age, gender, ISS, and mechanism of injury were used to form the TXA and control groups. Matching of each patient for the TXA group and control group was performed within the trauma registry of each center involved. We conducted chi-square analyses to identify whether there was a difference in the mortality at 24 hours, 48 hours, and 28 days between the TXA and control groups. Independent T-tests were conducted to identify whether there were differences of continuous variables (e.g., age) between the TXA and control groups.

Wilcoxon rank-sum tests were conducted to identify whether the median of some continuous variables (e.g., hospital LOS) was different between the TXA and control groups. Based on the original study design, we conducted three subgroup analyses to assess patient outcomes including (1) those who received one dose of TXA in comparison to two doses of TXA; (2) those who sustained significant blood loss (≥10 units of total blood products transfused) and those who did not sustain significant blood loss (<10 units of total blood products transfused), similar to the subanalysis performed in the Military Application of Tranexamic Acid in Trauma Emergency Resuscitation (MATTERs) study;22 (3) those who were severely injury (ISS ≥16) and those who were less severely injured (ISS <16).

The original sample-size calculation was based on the published results using 48-hour mortality as the primary outcome. Morrison and colleagues suggested that the TXA 48-hour mortality rates were 11.3% and 18.9% for TXA and control.22 Controlling for the type I error rate of 0.05, a sample size of 369 patients in each group would achieve a statistical power of 0.80.