|Home | About | Journals | Submit | Contact Us | Français|
Hemorrhage and trauma induced coagulopathy remain major drivers of early preventable mortality in military and civilian trauma. Interest in the use of prehospital plasma in hemorrhaging patients as a primary resuscitation agent has grown recently. Trauma center-based damage control resuscitation using early and aggressive plasma transfusion has consistently demonstrated improved outcomes in hemorrhaging patients. Additionally, plasma has been shown to have several favorable immunomodulatory effects. Preliminary evidence with prehospital plasma transfusion has demonstrated feasibility and improved short-term outcomes. Applying state-of-the-art resuscitation strategies to the civilian prehospital arena is compelling. We describe here the rationale, design, and challenges of the Prehospital Air Medical Plasma (PAMPer) trial. The primary objective is to determine the effect of prehospital plasma transfusion during air medical transport on 30-day mortality in patients at risk for traumatic hemorrhage. This study is a multicenter cluster randomized clinical trial. The trial will enroll trauma patients with profound hypotension (SBP ≤ 70 mmHg) or hypotension (SBP 71–90 mmHg) and tachycardia (HR ≥ 108 bpm) from six level I trauma center air medical transport programs. The trial will also explore the effects of prehospital plasma transfusion on the coagulation and inflammatory response following injury. The trial will be conducted under exception for informed consent for emergency research with an investigational new drug approval from the U.S. Food and Drug Administration utilizing a multipronged community consultation process. It is one of three ongoing Department of Defense-funded trials aimed at expanding our understanding of the optimal therapeutic approaches to coagulopathy in the hemorrhaging trauma patient.
Hemorrhage and coagulopathy remain major drivers of preventable early death following injury.1–4 Recent studies focus on early resuscitation to reduce this burden.5–13 With the increasing recognition that trauma-induced coagulopathy (TIC) is an early phenomenon with a significant deleterious impact on outcomes, focus has shifted to directly addressing this entity.14 High ratio of plasma to red blood cell (RBC) transfusion is associated with improved outcomes in both military and civilian settings.5,7,12,15,16 Given the encouraging results from the military and trauma center resuscitation strategies, applying the lessons learned to the civilian prehospital setting is appealing.
To this end, the U.S. Department of Defense (DoD) issued the Prehospital Use of Plasma for Traumatic Hemorrhage program announcement, seeking to fund three awards to study the use of plasma for traumatic hemorrhage in the prehospital environment.17 While the use of early plasma has been incorporated into U.S. military damage control resuscitation guidelines as well as many civilian trauma centers, uncertainty remains about the effectiveness of plasma in the pre-hospital setting.18 The Prehospital Air Medical Plasma (PAMPer) trial seeks to address this question.
As the subject of intense recent study, our understanding of TIC has evolved significantly.2,19–21 Evidence suggests that 25% of severely injured patients are coagulopathic at admission, independent of hemodilution.22,23 Further evidence demonstrates coagulopathy develops within minutes of injury. Patients that develop TIC consistently have worse outcomes, with documented increases in morbidity and mortality.2,19,24,25
Hypoperfusion in trauma and hemorrhage results in excessive activation of protein C, which leads to rapid depletion of multiple hemostatic factors and hyperfibrinolysis.26,27 Hyperfibrinolysis is a strong predictor of massive transfusion of mortality.23,28 Hyperfibrinolysis also has emerged as a promising therapeutic target, based on decreased mortality in patients receiving anti-fibrinolytics in the CRASH-II trial.29 Our group is also leading a trial to examine the effects of prehospital administration of tranexamic acid in trauma patients.30 Finally, the identification of early platelet dysfunction is a key new area of investigation in TIC research. Platelets play a key role not only in the hemostatic response following injury but also as inflammatory sentinels, serving to activate other immune cell types at the site of injury.31–33
Thus, TIC involves a complex cascade of events related to direct tissue injury, hypoperfusion, and inflammation that results in systemic activation of protein C, hyperfibrinolysis, endothelial and platelet dysfunction, complement activation, and release of inflammatory mediators.34–39
Analysis of U.S. combat casualties reveals that non-compressible hemorrhage underlies the majority of potentially preventable mortality.40 The military responded by devising a resuscitation strategy to replace blood components in a fashion that attempts to reconstitute whole blood. Beginning with the landmark paper by Borgman and colleagues, several retrospective studies in military and civilian populations demonstrate increasing ratios of plasma and platelets to RBCs are associated with improved outcomes.5,7,12,16 This has been confirmed in the prospective PROMMTT cohort study, where increased ratios of plasma to RBCs are associated with increased survival over the first 24 hours.41 Based on these studies, most trauma centers in the United States aim to provide plasma to RBC ratios between 1:1 and 1:2.
Plasma has garnered significant interest recently for use in prehospital resuscitation. Prehospital plasma transfusion replaces critical coagulation factors deficiencies at the point of injury, which can occur as early as 15 minutes from injury.40,42,43 Plasma also represents an ideal volume expander, which remains in the intravascular space. Plasma contains several immunologically active and potentially cytoprotective proteins.44 Further, plasma prevents endothelial glycocalyx degradation and reduces endothelial permeability leading to an improved inflammatory response to injury compared to conventional isotonic crystalloids.40,42 Early experience with prehospital plasma resuscitation is favorable, demonstrating feasibility and improvement in short-term outcomes such as transfusion requirements and TIC parameters.45–47
In response to the DoD request for proposals for trials to study the effectiveness of prehospital plasma transfusion in traumatic shock, we designed the PAMPer trial as a multicenter randomized clinical trial.48 The primary objective of this trial is to determine the effect of transfusion of 2 units of AB plasma during prehospital air medical transport on 30-day mortality in trauma patients with hemorrhagic shock as compared to standard air medical care. We hypothesize that prehospital plasma transfusion will reduce mortality as well as reduce in-hospital transfusion requirements and improve measures of coagulopathy. The specific aims for the PAMPer trial are included in Table 1.
PAMPer is designed as a multicenter prospective open-label cluster randomized trial. Participating sites include the University of Pittsburgh (coordinating site), Case Western University, University of Louisville, University of Texas Southwestern, University of Tennessee, and Vanderbilt University. All sites are large academic level I trauma centers with active air medical helicopter transport programs.
Eligible subjects include blunt or penetrating trauma patients undergoing air medical transport to one of the participating centers with 1) systolic blood pressure (SBP) ≤ 70 mmHg, or 2) SBP 71–90 mmHg with heart rate (HR) ≥ 108 bpm (SBP and HR criteria need not be simultaneous). Vital sign inclusion criteria are met if documented at the scene, any time during transport, or at an outside referring facility while awaiting transport. We selected these inclusion criteria to identify patients at highest risk for hemorrhagic shock and thus the greatest chance of benefit from the intervention. Exclusion criteria include age <18 or >90 years old, inability to obtain intravenous or interosseous access, isolated fall from standing mechanisms, radiographically documented cervical spinal cord injury with motor deficit, prisoner, pregnancy, cardiac arrest >5 minutes without return of vital signs, penetrating cranial injury, traumatic brain injury with brain matter exposed, isolated drowning or hanging victims, isolated burn injury with >20% estimated total body surface area, referral from facility for an in-patient, or wearing an opt-out bracelet (see below).
The intervention arm of the trial includes provision of standard air medical care plus two units of AB thawed plasma between 0 and 5 days old. Plasma will be provided to air medical bases participating in the trial by each center's blood bank and will be stored at the air base and on the helicopter according to established standards. The respective blood banks will regularly exchange any unused plasma to allow use of plasma by the blood bank in-hospital while ensuring the air bases remain stocked with plasma for the study.
Given the challenges in providing and maintaining plasma in the prehospital environment, and the inability to blind blood product transfusion, a single-stage cluster randomization scheme was utilized. Between two and six air medical bases at each participating center will be block randomized to the intervention or control arm for 1-month periods. Block sizes will vary randomly over the entire trial enrollment period; the randomization scheme was generated by a computer random number generator. Thus, each participating air base will be the unit of randomization to either a plasma (intervention) month or control month over the study duration. Due to the nature of the intervention, blinding will not be performed for patients, prehospital providers, trauma center staff, or enrolling research staff as to the air base trial arm assignment or treatment received by individual subjects in the study. Separate outcome assessors will not be informed of the study intervention received and will be blinded where possible.
Subjects who meet inclusion and exclusion criteria and are transported by a helicopter assigned to a plasma intervention month will have the initiation of 2 units of AB thawed plasma transfused through an intravenous or interosseous line (Figure 1). In the event that the plasma transfusion is complete prior to arrival at the participating trauma center, a goal-directed resuscitation protocol will be followed in addition to standard air medical care.6 Subjects with persistent hypotension (SBP < 90 mmHg) after plasma transfusion will receive either a 500-mL crystalloid bolus or RBC transfusion if the capability exists under the air medical provider's existing protocols. When hypotension is resolved or absent, crystalloid infusion to maintain intravenous line patency or heparin–saline lock will be provided only. In the event the plasma transfusion is not completed prior to arrival at the participating center, the transfusion of a total 2 units of plasma will be completed in the trauma resuscitation area, with further resuscitation at the discretion of the attending trauma surgeon.
Subjects who meet inclusion and exclusion criteria and are transported by a helicopter assigned to a control month will undergo standard air medical care with goal-directed resuscitation described above. Standard air medical care will be that under each air medical providers' existing operating procedures. Three air medical providers included in the trial carry and administer up to 2 units of O-negative RBCs for patients with persistent hypotension following 1–2 L of crystalloid or clinical evidence of ongoing hemorrhage. A sham plasma bag will be placed with each control patient by the air medical providers to maintain focus on the study and minimize potential hospital provider bias. The goal was to generally reduce treatment bias by hospital-based treatment team with the appearance of an empty plasma bag to casual observation; however, if questions arise regarding complications potentially related to blood product administration that require immediate attention, careful inspection will allow determination by the treating providers whether the patients received the plasma intervention or not.
Upon arrival at the participating site, 24-hour available research staff will enroll subjects through a web-based secure program built for the trial. No further study intervention will be performed outside of completing transfusion of the 2 units of AB plasma if not completed en route. The participating sites will continue care of the subjects according to local protocols and standards of care. All sites have damage control resuscitation protocols, including aggressive blood product administration based on best practices established in the literature, which should reduce significant variation in early care.
To address the aims of the study, research staff will collect blood for laboratory analysis and outcome data at several time points. Upon arrival, blood will be drawn for laboratory analysis. Blood will also be collected at 24 and 72 hours. Laboratory analysis includes rapid thromboelastography (TEG), conventional coagulation parameters (prothrombin time, partial thromboplastin time, international normalized ratio), base deficit, lactate, complete blood count, and D-dimer. As part of the investigation into the inflammatory response, serum IL-6, thrombomodulin, and protein C-levels will be determined using standard immunoassay techniques. Mortality outcomes will be collected at 24 hours, 30 days, and in-hospital. Resuscitation requirements of packed RBCs, plasma, platelets, and crystalloid will be collected at 3, 6, 12, 18, and 24 hours. Specific adverse events to be collected in-hospital include multiple organ failure (MOF), nosocomial infection (NI), acute lung injury (ALI), and transfusion acute lung injury (TRALI).
Sample size was determined using the primary outcome of 30-day mortality. We estimated a 22% baseline mortality based on mortality for patients requiring 4 units of blood transfusion in the first 6 hours from injury in the Inflammation and the Host Response to Injury study conducted at eight large academic level I trauma centers in the United States, which included two participating sites for this trial.49 We sought to detect a 14% absolute reduction in 30-day mortality again based on mortality in subjects receiving high plasma to RBC transfusion in the Inflammation and the Host Response to Injury study. Using a two-sided alpha of 0.05 and power of 88%, 144 subjects are required per arm. This sample size was corrected by a factor of 1.75 to account for the clustered randomization design and 5% missing outcome data for a final sample size of 265 subjects per arm with 530 subjects total. Based on the expected volume of eligible patients across the participating sites, recruitment will occur over a 4-year period to achieve our sample size. All analyses will be conducted as intention to treat and site stratification will be accounted for in statistical tests. Generalized estimating equations regression will be used to adjust for clustering as well as prehospital transport time and prehospital RBC transfusion in outcome analysis. Several predefined exploratory subgroup analyses are planned in subjects based on mechanism of injury, requirement for massive transfusion, and preexisting coagulopathies.
This trial will enroll subjects under exception from inform consent for emergency research as outlined under U.S. Federal regulation 21 CFR 50.24.50 Under this regulation, any study that will enroll under an exception from informed consent must meet the conditions in Table 2.51 For studies pursuing an exception from inform consent for emergency research, the institutional review board (IRB) must determine whether the proposed intervention and study falls under the regulation of the FDA. Since our intervention falls within the jurisdiction of the FDA, we are required to obtain an investigational new drug application (IND# 15117). This process is required due to the waiver of consent even though plasma is an FDA-regulated and American Association of Blood Bankers (AABB)-approved therapeutic.
The exception from informed consent for emergency research also requires a community consultation process at each participating site. This process consists of two parts: 1) public notification, and 2) community consultation. For this study, a multipronged approach was utilized and has been completed. The notification process included distribution of information regarding the trial in several mediums. A website was created within the existing acute care research website for the University of Pittsburgh (acutecareresearch.org). Each participating center also developed an informational website for the trial. Bumper ads on Pittsburgh Port Authority buses were designed to direct people to the website. Fliers were distributed throughout area hospital waiting rooms and local community center boards. In-person presentations were conducted for several local-area EMS, fire, and police agencies regarding the study. Additionally, an informational video is available on YouTube (www.youtube.com/watch?v=fWgC1nXwGwE).
The consultation process involves monitoring traffic and hits on the above websites. Additionally, the website allows email communication with the study coordinators either anonymously or with return contact information. A telephone survey was conducted by an independent company with random digit dialing for 500 households in each study region. The survey provides information regarding demographics, social economic status, willingness to be enrolled in the trial, and potential concerns. Periodic sampling of the experience of enrolled subjects and family were recorded.
Finally, opt-out bracelets are available to any community member by phone or by email. The bracelet is to be worn at all times by the individual who prospectively does not wish to be enrolled in the case of severe injury, and air medical providers have been trained to look for these prior to enrolling the patient. For patients who are enrolled, informed consent will be obtained to use the patient's data from the patient or legally authorized representative as soon as feasible after in the intervention, as required for exception from informed consent. Although this process is demanding, it is necessary to assure ethically responsible and high-quality emergency research trials, while maintaining public trust and confidence.
As with any trial conducted in the course of clinical care, particularly under emergency conditions, several potential pitfalls and limitations may be anticipated. Given the nature of the intervention, blinding of patients and providers is not feasible. This limitation exposes the trial to potential bias and may influence subsequent treatment decisions by providers. We have attempted to mitigate this by blinding outcome assessors as well as selecting primary and secondary outcomes that are objective and robust to bias.
Intervention nonadherence is a potential source of bias. We anticipate the most common nonadherence will be subjects not receiving the intervention when eligible (drop-out), and expect a 3–5% rate. We also expect a small proportion (1% or less) of noneligible subjects receiving the intervention (drop-in). Subjects in these circumstances will be analyzed as intention-to-treat in the assigned treatment group. Further, we anticipate a 3% missing data rate in the primary outcome based on the objective and easily collected nature of mortality. We have adjusted the sample size to account for up to 5% missing data, and have planned sensitivity analyses using multiple imputation. We have designated a 15% rate of missing primary outcome data as unacceptable for analysis and in this case would report the outcome as descriptive.
Finally, the use of a cluster design significantly inflates the necessary sample size, and feasibility of adequate enrollment requires careful consideration. In the event that 30-day mortality is not significantly reduced by the intervention, we strongly anticipate other significant and clinically relevant advantages will be seen in the plasma arm when analyzing secondary outcomes.
The aims of this trial have implications for the care of both civilian and military trauma patients, and allow evidence-based decisions regarding the incorporation of plasma into prehospital resuscitation protocols. Furthermore, this trial will lend insight into the inflammatory response after injury and resuscitation. This trial is one of three awards funded by the DoD to investigate prehospital plasma transfusion in traumatic shock. The Denver group (PI: Gene Moore, MD) has undertaken the Control of Major Bleeding after Trauma (COMBAT) trial with ground ambulance plasma administration, looking at the effect on TIC, transfusion requirements, metabolic recovery, organ failure, and mortality.52 The Prehospital Use of Plasma for Traumatic Hemorrhage (PUPTH) trial through Virginia Commonwealth University (PI: Bruce Spiess, MD) will also examine the impact of using thawed prehospital plasma on mortality and coagulopathy.53
Under the auspices and collaboration of the Department of Defense and the National Heart Lung and Blood Institute (NHLBI), a Transagency Collaboration for Trauma Induced Coagulopathy (TACTIC) consortium was initiated and funded, which is now beginning the difficult task of deciphering the mechanisms and driving factors that promote TIC following injury. Most of the inclusion criteria, safety events, laboratory analysis, and outcomes of the three DoD-funded pre-hospital plasma trials are harmonized to allow pooled analysis of enrolled subjects and promote generalizability and sharing of resources between trial infrastructures. This approach enhances the potential for reaching meaningful conclusion from the data, and should be considered for trials conducted in the future. These efforts will contribute tremendously to our understanding of the mechanism and optimal therapeutic approaches to coagulopathy in the hemorrhaging trauma patient.
This trial is supported by the following grant: Department of Defense W81XWH 12-2-0023.
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.