Association of Prehospital Plasma Transfusion With Survival in Trauma Patients With Hemorrhagic Shock When Transport Times Are Longer Than 20 Minutes: A Post Hoc Analysis of the PAMPer and COMBAT Clinical Trials

Key Points

Question  Is prehospital plasma administration more beneficial when patient transport times are longer?

Findings  This post hoc analysis was performed using harmonized data from 2 randomized clinical trials, Control of Major Bleeding After Trauma and Prehospital Air Medical Plasma, which included 626 patients with trauma and hemorrhagic shock. Patients who received prehospital plasma transfusion had significantly reduced 28-day mortality compared with standard care when prehospital transport times were longer than 20 minutes.

Meaning  Prehospital plasma administration is associated with reduced mortality in patients with trauma and significant hemorrhage when transport times are prolonged.

Importance  Both military and civilian clinical practice guidelines include early plasma transfusion to achieve a plasma to red cell ratio approaching 1:1 to 1:2. However, it was not known how early plasma should be given for optimal benefit. Two recent randomized clinical trials were published, with apparently contradictory results. The Prehospital Air Medical Plasma (PAMPer) clinical trial showed a nearly 30% reduction in mortality with plasma transfusion in the prehospital environment, while the Control of Major Bleeding After Trauma (COMBAT) clinical trial showed no survival improvement.

Objective  To facilitate a post hoc combined analysis of the COMBAT and PAMPer trials to examine questions that could not be answered by either clinical trial alone. We hypothesized that prehospital transport time influenced the effects of prehospital plasma on 28-day mortality.

Design, Setting, and Participants  A total of 626 patients in the 2 clinical trials were included. Patients with trauma and hemorrhagic shock were randomly assigned to receive either standard care or 2 U of thawed plasma followed by standard care in the prehospital environment. Data analysis was performed between September 2018 and January 2019.

Interventions  Prehospital transfusion of 2 U of plasma compared with crystalloid-based resuscitation.

Main Outcomes and Measures  The main outcome was 28-day mortality.

Results  In this post hoc analysis of 626 patients (467 men [74.6%] and 159 women [25.4%]; median [interquartile range] age, 42 [27-57] years) who had trauma with hemorrhagic shock, a Cox regression analysis showed a significant overall survival benefit for plasma (hazard ratio [HR], 0.65; 95% CI, 0.47-0.90; P = .01) after adjustment for injury severity, age, and clinical trial cohort (COMBAT or PAMPer). A significant association with prehospital transport time was detected (from arrival on scene to arrival at the trauma center). Increased mortality was observed in patients in the standard care group when prehospital transport was longer than 20 minutes (HR, 2.12; 95% CI, 1.05-4.30; P = .04), while increased mortality was not observed in patients in the prehospital plasma group (HR, 0.78; 95% CI, 0.40-1.51; P = .46). No serious adverse events were associated with prehospital plasma transfusion.

Conclusions and Relevance  These data suggest that prehospital plasma is associated with a survival benefit when transport times are longer than 20 minutes and that the benefit-risk ratio is favorable for use of prehospital plasma.

Trial Registration  ClinicalTrials.gov identifiers: NCT01838863 (COMBAT) and NCT01818427 (PAMPer)

Over the past 10 years, the critical role of initial blood component transfusion for resuscitation following severe trauma and hemorrhagic shock has been demonstrated, and early transfusion has been incorporated into military and civilian clinical practice guidelines.^1-3 In contrast to earlier approaches, which relied heavily on crystalloids and red blood cells (RBCs), the emphasis is to include plasma early to achieve a 1:1 to 1:2 plasma to RBC ratio.⁴ The survival benefit of early plasma is most evident among patients likely to die within the first 6 hours as a result of bleeding.^5-7 Studies conducted between 2015 and 2018 have demonstrated a survival benefit associated with initiating transfusions earlier, at the scene of injury or en route to a trauma center.^8-10 Consequently, a number of trauma systems have begun to incorporate RBCs, plasma, or whole blood in the prehospital setting.^8,9,11-19

Two prospective randomized studies of the prehospital administration of plasma were recently completed.^10,20 The US Department of Defense and the National Heart, Lung, and Blood Institute worked collaboratively by harmonizing the studies in terms of design and data collection and by sharing the samples and data.^21,22Quiz Ref ID Sperry et al¹⁰ conducted a multicenter study of more than 500 trauma patients with hemorrhagic shock who were transported by helicopter. Patients received standard care en route with or without the addition of 2 U of thawed plasma prior to other resuscitation measures. Prehospital administration of plasma resulted in a significantly lower 30-day mortality (23.2% vs 33.0%; P = .03) compared with the standard care group. Quiz Ref IDIn contrast, Moore et al²⁰ reported that 2 U of thawed plasma prior to other fluids during ground ambulance transport in a single-center clinical trial (with short transport times and immediate in-hospital access to blood components) did not improve survival. Recent commentaries have addressed the potential implications of these studies.^23-25

The reasons for these apparently contradictory results are not clear. One hypothesis is that the very short prehospital transport times in the ground ambulance study may have eliminated the potential for prehospital plasma to improve survival because in-hospital transfusion was not delayed significantly by transport. It was not possible to determine a time effect within either study independently, but analysis of the combined data from both studies offers the opportunity to examine this question. Therefore, we examined the combined data set to address the post hoc hypothesis that the benefits of prehospital administration of plasma are influenced by prehospital transport time.

This analysis brings together data from 2 previously published studies, the Control of Major Bleeding After Trauma (COMBAT) and the Prehospital Air Medical Plasma (PAMPer) clinical trials.^10,20,26,27 These clinical trials were harmonized in advance to enable a combined per-patient analysis to address questions that could not be answered by either trial individually. During protocol development, harmonization was performed to standardize as much as possible the 2 studies in the following key areas: (1) experimental treatment groups, (2) inclusion and exclusion criteria, (3) timing of blood samples, (4) monitoring of adverse events, (5) methods to account for patient transport time, and (6) data collection.

Data sets were developed for each study independently and provided to the data coordinating center of the closely aligned Trans-Agency Consortium for Trauma-Induced Coagulopathy (TACTIC).²² The TACTIC data coordinating center established the combined data set, ensured agreement of all data elements, and provided the combined data set for the present post hoc analysis, which was performed between September 2018 and January 2019.

Because of the pragmatic character of the clinical trials and requirements for rapid enrollment and randomization, the studies were exempted from the requirement for advanced written informed consent. Each individual clinical study protocol (COMBAT and PAMPer) was approved by its respective local institutional review board and by the Human Research Protections Office of the US Army Medical Research and Materiel Command. The protocols are available in Supplement 1. The requirement to obtain informed consent for emergency research was waived in accordance with Code of Federal Regulations Title 21, Part 50—Protection of Human Subjects, Subpart B—Informed Consent in Human Subjects and SEC 50.24—Exception from Informed Consent Requirement for Emergency Research.

COMBAT was a pragmatic randomized placebo-controlled single-center clinical trial. Eligible patients were assessed and enrolled at the scene according to the harmonized inclusion and exclusion criteria (eTables 1 and 2 in Supplement 2). Patients were transported by ground ambulance directly from the scene to an urban level 1 trauma center with blood components immediately available in the emergency department (ED).

Patients enrolled in the COMBAT clinical trial were administered either 2 U of thawed AB plasma (universal donor plasma of approximately 250 mL each) followed by standard care or standard care with crystalloid en route. Plasma was administered intravenously by paramedics in the ambulance before other resuscitative fluids were initiated. Plasma transfusion was continued into the hospital setting if necessary to complete the 2 U. Standard care was goal-directed crystalloid resuscitation using 0.9% saline. Time of arrival on scene (AOS) and time of arrival at the trauma center were recorded by ambulance staff. Randomization and enrollment were performed at the level of the ambulance.^20,26

The PAMPer clinical trial was a pragmatic multicenter cluster-randomized clinical trial involving injured patients who were transported by air medical transport to a level 1 trauma center, either directly from the scene or from a referring hospital. Eligible patients were assessed and enrolled at the scene according to the harmonized inclusion and exclusion criteria (eTables 1 and 2 in Supplement 2).

Patients enrolled in PAMPer received 2 U of either group AB or group A with a low anti-B antibody titer (<1:100) thawed plasma followed by standard care, or standard care. Plasma was administered by paramedics prior to other resuscitation fluids. Both units of the prehospital-initiated plasma were infused to completion even if the infusion was still ongoing at the time of arrival at the trauma center. In cases in which completion of the infusion of the 2 U of plasma occurred during flight, standard trauma resuscitation (as defined by the local protocol) resumed until arrival at the trauma center. Standard care consisted of goal-directed crystalloid-based resuscitation on the basis of hemodynamic status for air transport teams at 14 of the 27 participating air medical bases. Air transport teams at the 13 other participating air medical bases also carried 2 U of universal donor RBC on all flights. If a patient remained hypotensive after the plasma infusion or had obvious bleeding, transfusion of RBC then proceeded according to the local protocol. Following RBC transfusion, these teams reverted to crystalloid-based resuscitation. Time of AOS and time of arrival at the trauma center were recorded by helicopter staff. Randomization was at the level of the air medical base.^10,27 Interventions for each of the studies are summarized (eTable 3 in Supplement 2).

The primary outcome measure was 28-day mortality. Secondary outcomes included 24-hour mortality, volumes of in-hospital blood components administered within 6 and 24 hours, ventilator-free days among patients alive at 28 days, intensive care unit–free days among patients alive at 28 days, and international normalized ratio.

Follow-up times were prespecified in this study as 28 days or 24 hours from randomization (or AOS) until death or censoring on the 28th day or 24th hour after AOS. Prehospital transport time was defined as time in minutes from ambulance AOS to arrival at the ED of the trauma center. The prehospital transport time was a priori defined as shorter or longer transport time if prehospital transport time was within 20 minutes or longer than 20 minutes, respectively. All efficacy analyses were carried out in the intention-to-treat randomized patients. A multivariate analysis of survival was performed with the use of a Cox proportional hazards model (for computing hazard ratios [HRs]) to evaluate the treatment effect (plasma vs standard care) and time effect (longer vs shorter), with adjustment for stratification factors and other possible confounding factors (age, injury severity score [ISS], and clinical trial cohort in overall models). The Kolmogorov-type supremum test was used for the Cox proportional hazards assumption. Cohort was included as a random effect because of the heterogeneity inherent in the 2 cohorts. Logistic regression models (for computing the odds ratios [ORs]) were used for likelihood of mortality.

Descriptive statistics characterized the demographics and injuries of the patients and outcomes of interest. Categorical variables were presented as frequencies and percentages and tested using a χ² test. Continuous variables were expressed as means and SDs or medians and interquartile ranges (IQRs) and were tested using the t test or Mann-Whitney test as appropriate. Statistical significance was determined at the P < .05 level (2-sided). All data were analyzed using SAS, version 9.4, and JMP 13 software (SAS Institute Inc).

We reviewed 705 patients who were randomly assigned to either the standard care group or the plasma group (Figure 1). A total of 626 patients (467 men [74.6%] and 159 women [25.4%]; median [IQR] age, 42 [27-57] years) met inclusion criteria for the primary outcome (Figure 1). Of those, 125 patients were reported for the COMBAT clinical trial and 501 were reported for the PAMPer clinical trial.^10,20 Among the 2 study cohorts, median (IQR) prehospital transport time was longer in the PAMPer study compared with the COMBAT study (41 [33-52] vs 18 [15-22] minutes, respectively; P < .001), but there was overlap between the 2 studies (eFigures 1, 2, and 3 in Supplement 2).

Randomization and harmonization procedures resulted in similar patients being enrolled in the plasma or standard care groups. Patient characteristics are described in Table 1. There were no significant differences observed in any demographic or injury characteristic, and there were no differences observed in baseline physiological status (heart rate and systolic blood pressure). Median transport times were also similar between study groups.

Quiz Ref IDThe 28-day mortality was lower in the plasma group (61 of 297 patients [20.5%]) than in the standard care group (94 of 329 patients [28.6%]) (P = .02) (Table 2 and Figure 2). The HR generated by a Cox regression model adjusted for age, injury severity, and trial cohort (COMBAT or PAMPer) indicated lower mortality in the plasma group (HR, 0.65; 95% CI, 0.47-0.90; P = .01) (Table 3). Quiz Ref IDA similar pattern was observed for 24-hour mortality (HR, 0.62; 95% CI, 0.42-0.93; P = .02) (Table 3). Most deaths in both groups occurred within the first 6 hours after injury (Figure 2).

A Cox regression model showed that, in addition to treatment group, survival was influenced by ISS, age, and prehospital transport time. Sensitivity analysis revealed that a change in response was evident for prehospital times longer than 20 minutes (P = .003 vs P = .007 for 17 minutes, P = .006 for 22 minutes, P = .01 for 25 minutes, and P = .02 for 30 minutes). Transport time (≤20 minutes vs >20 minutes) was not associated with survival when examined across treatment groups (HR, 1.37; 95% CI, 0.85-2.21; P = .20) (Figure 2). However, stratified analysis revealed that in patients who received standard care, rate and likelihood of mortality were significantly increased by 2-fold with transport times greater than 20 minutes (HR, 2.12; 95% CI, 1.05-4.30; P = .04) (Table 3 and Figure 2). Among patients who received prehospital plasma, this association with transport time was eliminated (HR, 0.78; 95% CI, 0.40-1.51; P = .46) (Table 3 and Figure 2). Among patients with short transport times (≤20 minutes), survival in the plasma group and the standard care group did not differ (HR, 1.71; 95% CI, 0.70-4.16; P = .24) (Table 3 and Figure 2). Quiz Ref IDAmong patients with longer transport times (>20 minutes), survival was improved in the plasma group (HR, 0.56; 95% CI, 0.40-0.80; P = .001) (Table 3 and Figure 2).

Patients who received prehospital plasma were 47% less likely to present to the ED with coagulopathy (international normalized ratio >1.3) compared with those who received standard care (OR, 0.53; 95% CI, 0.35-0.80; P = .002) (eFigure 4 in Supplement 2), and this association was isolated to the group with transport times longer than 20 minutes. Among patients with transport times of 20 minutes or less, in-hospital transfusion requirements did not differ for RBC, fresh frozen plasma, and platelets in the first 6 hours after ED admission, while patients who received plasma during longer transports required less in-hospital transfusion, with median (IQR) in-hospital transfusion requirements of 5 (2-10) vs 2 (2-4) U of plasma (P < .001), 5 (3-10) vs 4 (2-8) U of RBC (P = .05), and 2 (1-3) vs 1 (1-2) U of platelets (P = .04) at 6 hours (Table 2). Similar results were found in the first 24 hours after ED admission (Table 2). Total plasma requirements (including prehospital plasma) did not differ between groups (data not shown). Intensive care unit–free days among patients alive at 28 days did not differ between groups, while ventilator-free days were slightly lower in the plasma group (Table 2). Secondary outcomes based on transport time are shown in eTable 4 in Supplement 2.

Prehospital administration of plasma was associated with significantly reduced 24-hour and 28-day mortality compared with standard care in this harmonized data set (Figure 2). This finding is consistent with that reported for the PAMPer clinical trial but not for the COMBAT clinical trial.^10,20 The ability to observe this overall association in the harmonized data set may have been owing to the larger overall number of patients included. This association appears to be robust since, even after adjustment for clinical trial cohort, age, and injury severity, the HR was 0.65 (Table 3). This finding is consistent with previous observations that the survival benefit of early in-hospital plasma transfusion is most substantial among patients likely to die as a result of bleeding within the first 6 hours of injury.^5-7 We also found that transport times longer than 20 minutes were associated with increased mortality in the standard care group and that this increase in mortality was mitigated when prehospital plasma was administered. The present findings suggest that prehospital plasma administration provides a benefit beyond that of a balanced in-hospital transfusion regimen, as was practiced at all involved centers in the COMBAT and PAMPer clinical trials.^10,20

In a recently published meta-analysis, it was suggested that the case for prehospital plasma could not yet be made in light of the differing results of the COMBAT and PAMPer studies.²⁴ We were able to adjust for confounding factors and to specifically address the potential effects of prehospital transport time. In the present analysis, the median prehospital transport time was 38 minutes and represented a broad range (Table 1; eFigure 1, eFigure 2, and eFigure 3 in Supplement 2). A survival advantage associated with prehospital plasma was observed in the PAMPer trial but not in the COMBAT trial.^10,20 The reason for these apparently contradictory results are not clear. One difference between the 2 primary studies was that the median prehospital transport times were substantially different (18 minutes in the COMBAT study vs 41 minutes in the PAMPer study). Considering the detrimental effects of transfusion delays on survival,^9,28 we hypothesized that the very short prehospital transport time in the COMBAT trial may have eliminated the potential for prehospital plasma to improve survival because in-hospital transfusion was not delayed to a degree sufficient to influence mortality. We found that prehospital transport time influenced the response to prehospital plasma. Prehospital transport times longer than 20 minutes were associated with increased mortality in patients who received standard care (Figure 2). In contrast, the increased mortality associated with longer transport times was eliminated in patients who received prehospital plasma (Figure 2).

In a study of 502 casualties evacuated by US military MEDEVAC in Afghanistan from 2012 to 2015, the initiation of transfusion within 15 minutes of MEDEVAC rescue (median time of MEDEVAC rescue, 29 minutes after injury) was associated with improved survival (mortality HR, 0.17), while delays beyond that eliminated the association.⁹ Another report found that early in-hospital delays in the initiation of transfusion were associated with progressively increasing mortality rates.²⁸ While prehospital transport time is a more available measure, the more pathophysiologically relevant measure is the time from injury to transfusion. In the present study, the time of injury was available only in a subset of patients. The median time from injury to AOS was 20 minutes. Extrapolating this to the observed dichotomy between transport times longer or shorter than 20 minutes, it may be estimated that the benefit associated with prehospital plasma was most evident in patients who could not be delivered for in-hospital transfusion within approximately 40 (20 plus 20) minutes of injury. This is similar to the total time to transfusion of 36 minutes reported by Shackelford et al,⁹ when time from injury to MEDEVAC rescue is included. All centers involved in COMBAT and PAMPer had blood products readily available in the ED, likely minimizing possible in-hospital delays. However, time from ED arrival to initiation of transfusion must be considered in estimating transfusion delays and may differ based on local availability.

The finding that transport times longer than 20 minutes were associated with increased mortality in the standard care group emphasizes the importance of minimizing time to definitive care, as recently demonstrated in a national database analysis.²⁹ The importance of rapid hemostasis must also be recognized, and potential delays in getting to an operating room for surgical hemostasis could be a more significant factor than the time to transfusion.

Other factors may explain the observed differences in survival. Most patients with prolonged prehospital transport times were transported by helicopter. It has been reported that helicopter transport is associated with a survival advantage that may be a result of the higher level of training among helicopter medical crews.^30-32 In the present study, participating ground ambulance crews included paramedics. In addition, overall survival was better with ground transport, although this occurred with shorter transport times. Therefore, a difference in personnel training is not likely to account for the time-related differences observed. The association with transport time may also be explained by injury severity. The ISS in the COMBAT trial, which included most patients with short transport times, was lower than that in the PAMPer trial. However, the differential association of prehospital transport time in patients who did or did not receive prehospital plasma remained significant even after adjusting for ISS in the regression model (Table 3).

Among patients with longer transport times, those who received prehospital plasma had lower early transfusion requirements (Table 2). Reduced transfusion requirements suggest improved hemodynamic stability among patients who received prehospital plasma. These patients also had improved international normalized ratios (Table 2; eTable 4 and eFigure 4 in Supplement 2). Plasma transfusion mitigates the coagulopathy that can complicate traumatic hemorrhage and has also been reported to improve inflammatory response after injury, reduce permeability of endothelial cells, reduce gut permeability, and mitigate metabolic derangements after trauma and hemorrhagic shock.^17,33-40 Therefore, reduced transfusion requirements may reflect improved hemostasis, improved endothelial integrity, or a more favorable inflammatory status.

It is important that no significant differences in safety outcomes and adverse events between the plasma and standard care groups were previously reported for the individual studies.^10,20 The lack of differences in intensive care unit–free days and the small difference in ventilator-free days in the present analysis are consistent with these observations. This suggests that the benefit-risk ratio is favorable for the prehospital administration of plasma in cases in which there is a doubt about how rapidly patients can be delivered for in-hospital transfusion. More logistically supportable products, such as dried plasma, are needed to enable the broader use of plasma in the prehospital setting.⁴¹

One limitation of the present analysis is the fact that the mode of transport differed in the 2 cohorts. Because other important aspects of the studies were harmonized and because there was some degree of overlap in transport times across the 2 studies, we believe that it is possible to draw generalizable conclusions regarding the influence of prehospital plasma and transport time on patient outcomes. Another limitation is that the pragmatic nature of the 2 studies precluded complete standardization of crystalloid type, specific plasma type, and the use of RBC. However, standardization did ensure that the common factor among all patients who received plasma was that they received 2 U before other standard care fluids. In addition, randomization ensured that the plasma treatment group was represented across all local variations in standard care. A third limitation is that the exact time from patient injury to administration of plasma could not be determined. Time of injury was documented for only a small subset of patients and, therefore, this analysis was not possible. Time to surgical hemostasis was also not recorded and could not be analyzed. Nonetheless, we believe that prehospital transport time is an important component of total prehospital time and the most relevant in terms of providing interventions such as plasma.

The present findings have important implications for the treatment of patients with traumatic hemorrhage when surgical care and in-hospital transfusion may be delayed, such as in military settings, in rural and remote trauma, and in civilian disaster scenarios. The benefit-risk ratio favors prehospital plasma, but logistical and cost constraints may limit feasibility. Thawed plasma is a viable option for helicopter ambulance systems but is more challenging for ground ambulances with short transport times.

Accepted for Publication: October 6, 2019.

Corresponding Author: Anthony E. Pusateri, PhD, US Army Institute of Surgical Research, 3698 Chambers Pass, Bldg 3611, Ste B, JBSA-Fort Sam Houston, San Antonio, TX 78234 (anthony.e.pusateri.civ@mail.mil).

Published Online: December 18, 2019. doi:10.1001/jamasurg.2019.5085

Author Contributions: Dr Pusateri had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Pusateri, E. Moore, H. Moore, Le, Guyette, Chapman, Brown, Miller, Claridge, Sperry.

Acquisition, analysis, or interpretation of data: Pusateri, Le, Guyette, Chapman, Sauaia, Ghasabyan, Chandler, McVaney, Brown, Daley, Harbrecht, Claridge, Phelan, Witham, Putnam, Sperry.

Drafting of the manuscript: Pusateri, E. Moore, H. Moore, Le, Sauaia, Ghasabyan, Miller, Sperry.

Critical revision of the manuscript for important intellectual content: Pusateri, E. Moore, Le, Guyette, Chapman, Chandler, McVaney, Brown, Daley, Harbrecht, Claridge, Phelan, Witham, Putnam, Sperry.

Statistical analysis: Pusateri, Le, Sauaia, Brown.

Obtained funding: E. Moore, Claridge, Sperry.

Administrative, technical, or material support: E. Moore, H. Moore, Le, Guyette, Chapman, Ghasabyan, Chandler, McVaney, Daley, Harbrecht, Phelan, Putnam, Sperry.

Supervision: Chapman, Ghasabyan, McVaney, Phelan, Witham, Sperry.

Conflict of Interest Disclosures: Dr H. Moore reported owning financial shares in Thrombo Therapeutics and providing consulting services to Instrumentation Laboratory outside the submitted work. Dr Guyette reported receiving grants from the Department of Defense during the conduct of the study. Dr Chapman reported receiving grants from the Department of Defense during the conduct of the study and serving on the board of Thrombo Therapeutics and providing consulting services to Haemonetics and Instrumentation Laboratory outside the submitted work. Dr Sauaia reported receiving grants from the National Institutes of Health and the Department of Defense during the conduct of the study. Dr McVaney reported receiving grants from the Department of Defense during the conduct of the study. Dr Daley reported receiving grants from the Department of Defense during the conduct of the study. Dr Harbrecht reported receiving grants from the National Institutes of Health and the Department of Defense during the conduct of the study and grants from the National Institutes of Health and the Department of Defense outside the submitted work. Dr Claridge reported receiving grants from the Department of Defense during the conduct of the study. Dr Sperry reported receiving grants from the Department of Defense during the conduct of the study. No other disclosures were reported.

Funding/Support: This work was supported by grants W81XWH-12-2-0028 (Control of Major Bleeding After Trauma clinical trial) and W81XWH-12-2-0023 (Prehospital Air Medical Plasma clinical trial) from the US Army Medical Research and Materiel Command and grant UM1 HL120877-05 (Trans-Agency Consortium for Trauma-Induced Coagulopathy) from the National Heart, Lung, and Blood Institute.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content of this article is the sole responsibility of the authors and does not represent the views or policies of the US Department of Defense, the US Army Institute of Surgical Research, the Denver Health Medical Center, the University of Pittsburgh Medical Center, the University of Tennessee Health Science Center, the Vanderbilt University Medical Center, the University of Louisville, Case Western Reserve University, the University of Texas Southwestern, the Texas Health Harris Methodist Hospital, or the National Institutes of Health.

Data Sharing Statement: See Supplement 3.

Additional Contributions: Duncan Donahue, PhD (Center for Environmental Health Research, Fort Detrick, MD), Charles Peterson, MD (Telemedicine and Advanced Technology Research Center, Fort Detrick, MD), Basil Golding, MD (Center for Biologics Evaluation and Research, US Food and Drug Administration, Silver Spring, MD), and Wilbur Malloy, MA (Congressionally Directed Medical Research Program, US Army Medical Research and Materiel Command, Fort Detrick, MD), assisted with the planning of the study harmonization; Stephen Wisniewski (University of Pittsburgh, Pittsburgh, PA) helped with database management; Andrei Kindzelski, MD, PhD (National Heart, Lung, and Blood Institute, Division of Blood Diseases and Resources, Bethesda, MD), assisted with the Trans-Agency Consortium for Trauma-Induced Coagulopathy program coordination; and Richard Weiskopf, MD (University of California, San Francisco, CA), contributed to the planning and critical review of this article. No compensation was received.