Secondary Logo

Journal Logo

Original Articles

Fresh Frozen Plasma Should be Given Earlier to Patients Requiring Massive Transfusion

Gonzalez, Ernest A. MD; Moore, Frederick A. MD; Holcomb, John B. MD; Miller, Charles C. PhD; Kozar, Rosemary A. MD, PhD; Todd, S Rob MD; Cocanour, Christine S. MD; Balldin, Bjorn C. MD; McKinley, Bruce A. PhD

Author Information
The Journal of Trauma: Injury, Infection, and Critical Care: January 2007 - Volume 62 - Issue 1 - p 112-119
doi: 10.1097/01.ta.0000250497.08101.8b
  • Free


Acidosis, hypothermia, and coagulopathy were identified more than 20 years ago as a deadly triad for patients presenting with exsanguinating hemorrhage.1 Widespread recognition provided to be a rationale for fundamental changes in the initial management of severely injured patients who present with exsanguinating hemorrhage. Regional trauma systems now triage these critically injured patients to Level I trauma centers, where prevention of hypothermia, damage control surgery, massive transfusion (MT) protocols, and early intensive care unit (ICU) triage for optimized resuscitation are standards of care. Despite these major advances, hemorrhage remains a leading cause of early death in both civilian trauma and military combat casualty care.2

Recognizing that acidosis, hypothermia, and coagulopathy are physiologic derangements likely to complicate early management of severely injured patients, we prospectively record variables describing these potential complications in high-risk patients as part of standardized ICU shock resuscitation. As part of ongoing performance improvement, we implemented a formal MT protocol in the late 1990s. Our protocol was to transfuse fresh frozen plasma (FFP) after the patient had received 6 units of packed red blood cells (PRBC).3 This delay in administering FFP is standard of care in many US trauma centers, and is based on the traditionally held belief that posttraumatic coagulopathy develops over time because of acidosis-, hypothermia-, and resuscitation-related hemodilution and consumption of factors.

Recent studies challenge this traditional thought and report that most severely injured patients are coagulopathic at emergency department (ED) admission and before aggressive resuscitation interventions.4–6 Hirshberg et al.7 described development of coagulopathy in trauma patients using a computer model with data from patients at Ben Taub Hospital (Houston, TX). The rationale for a computer model was their observation that previous hemodilution models used to develop recommendations for MT grossly underestimate clotting factor needs, and that MT protocols based on these earlier models were validated in the 1980s when whole blood transfusion (not PRBC and component therapy) was standard of care. In their analysis, they identified that (1) early prolongation of prothrombin time (PT) is the sentinel event; (2) early administration of FFP is key in preventing coagulopathy; and (3) the optimal replacement ratio of FFP:PRBC is 2:3. They recommend that 2 units FFP be given with the first unit of PRBC in patients who are at high risk of requiring MT. After this report and similar observations at our trauma center by intensive care fellows with previous military combat casualty experience, we undertook this study to ascertain whether our standard of care MT protocol should be changed. Based on recently published data, we hypothesize that our MT protocol does not adequately prevent or correct coagulopathy, and that early uncorrected coagulopathy is predictive of mortality.


High-risk patients who were admitted to the Shock Trauma ICU at Memorial Hermann Hospital (Level I regional trauma center serving southeast Texas and a population of ∼4 million) and who met specific criteria underwent a 24-hour standardized shock resuscitation process directed by computerized decision support. This protocol has been described previously.8–10 Data describing the patients’ clinical course and resuscitation process were obtained prospectively. Criteria for the resuscitation protocol were (1) major torso trauma, defined as injury of two or more abdominal organs, two or more long bone fractures, complex pelvic fracture, flail chest, or major vascular injury; (2) metabolic stress, defined as base deficit (BD) ≥ 6 mEq/L within 12 hours of hospital admission; and (3) anticipated transfusion requirement of ≥ 6 units PRBCs within 12 hours of hospital admission, or age ≥ 65 years with any two of the three previous criteria. Patients with these criteria who also incurred severe brain injury, defined as Glasgow Coma Scale score ≤ 8 in the ICU and abnormal brain computed tomography scan finding, were not resuscitated by this standardized protocol unless the patient’s brain injury was assessed by the attending neurosurgeon to be at low risk of worsening cerebral edema with crystalloid volume loading. At ICU admission, a pulmonary artery catheter with continuous cardiac output monitoring capability and an arterial catheter were placed. Hemoglobin concentration ([Hb]) was monitored at bedside using a point-of-care analyzer (HemoCue; HemoCue Inc, Lake Forest, Calif.). [Hb], cardiac index (CI), and pulmonary capillary wedge pressure (PCWP) were the key measurement variables that were used to guide protocol logic. The process involves maintenance of oxygen delivery index (DO2I) ≥ 500 mL/min-m2 (≥ 600 mL/min-m2 before January 2001) with interventions of PRBC if [Hb] < 10 g/dL and DO2I < 500; crystalloid fluid bolus (1 L LR) if [Hb] ≥ 10, PCWP < 15 mm Hg, and DO2I < 500; PCWP-CI optimization (“Starling curve”) if [Hb] ≥ 10, PCWP ≥ 15, and DO2I < 500; inotrope infusion (milrinone) if PCWP-CI optimized, [Hb] ≥ 10, PCWP ≥ 15, and DO2I < 500; and vasopressor infusion (norepinephrine) if inotrope infusion ongoing, PCWP-CI optimized, Hb ≥ 10, PCWP ≥ 15, DO2I < 500 and mean arterial pressure < 60. This standardized shock resuscitation protocol directs the above interventions during the first ICU day. Data describing acidosis, hypothermia, and coagulopathy were obtained prospectively as a part of this process. At the start of the shock resuscitation protocol, baseline body core temperature (T), arterial blood gas, and coagulation profile comprising PT, international normalized ratio (INR), platelet count ([plt]), partial thromboplastin time (PTT), and fibrinogen concentration ([fib]) were obtained and repeated every 4 hours for the duration of the 24 hour process. Additional data characterizing the pre-ICU course were recorded retrospectively, and these data were recorded in a Trauma Research Database. The Trauma Research Database is maintained with approval of the Committee for the Protection of Human Subjects (Institutional Review Board) of the University of Texas Health Science Center at Houston.

Although prevention and/or treatment of acidosis, hypothermia, and coagulopathy were recognized as an important adjunct of shock resuscitation during the development of that protocol, these aspects of care were not rigorously managed by computerized decision support. As specified by our Level I Trauma Center standards of care, documentation and prevention of hypothermia began by measuring T in the ED using a urinary catheter T sensor and, after initial trauma evaluation was complete, the exposed body was covered with warmed blankets and fluids, and blood products were infused via fluid warming devices (e.g., Level I fluid infusor). In the operating room (OR), forced warm air blankets (Bair Hugger, Arizant Healthcare Inc, Eden Prairie, MN) were applied, the head was covered with a heat reflecting cap, and mechanical ventilation was provided with warm (38° C), humidified air. Fluid warmers (e.g., Level I fluid infusor) were also used together with these interventions in the ICU as needed to normalize T. T data in the ICU were obtained prospectively using the pulmonary artery catheter. Acidosis was managed by resuscitation and mechanical ventilation. Patients requiring shock resuscitation were intubated and ventilated to normalize Paco2. Metabolic acidosis responded to resuscitation interventions and was not specifically treated with intravenous buffer administration. We did not advocate sodium bicarbonate therapy in the ED, OR, or ICU unless arterial pH < 7.20. Our response to coagulopathy in pre-ICU settings was according to an empiric MT protocol that was initiated by the trauma surgeon. After 6 units PRBC were transfused in ED, OR, or interventional radiology facilities, the hospital blood bank was notified. The blood bank then sent 6 units PRBC and 4 units FFP in an insulated container, and these components were transfused upon receipt and type/cross match check. After these components were given, additional containers were provided by the blood bank as needed with 6 units PRBC, 6 units FFP, and a number of platelet 6-packs equal to the number of 12 unit quantities of PRBC that had been transfused, and these components were transfused. When the patient arrived in the Shock Trauma ICU, this empiric protocol was stopped by the bedside ICU physician. Component therapy (FFP, plt, cryoprecipitate) were administered to correct abnormal PT and maintain [plt] ≥100 kcells/mm3. Before 2002, this was done at the discretion of the Shock Trauma ICU critical care team. Typically, FFP (2 units), plt (6 pack), and/or cryoprecipitate (10 pack) were given and coagulation measurements were rechecked. In 2002, we began development of a rule-based, data-driven protocol for coagulopathy prevention and correction in the shock Trauma ICU, and implemented this at bedside as a paper protocol to more rigorously control this process of care.6 For purposes of this protocol, we chose INR > 1.3 as a threshold for FFP administration to correct coagulopathy during shock resuscitation of the actively bleeding patient. For purposes of this study, we defined coagulopathy as INR >1.2, moderate coagulopathy as 1.4 ≤ INR ≤ 1.8, and severe coagulopathy as INR > 1.8.

During the 51 months ending January 2003, there were 200 shock resuscitation protocol patients, of which 97 received MT, defined as ≥10 units PRBC in the first 24 hospital hours. Data were extracted from the Trauma Research Database for this study cohort describing demographics, pre-ICU course, ICU resuscitation, and outcomes, with the focus on hypothermia, acidosis, and coagulopathy.

Data are presented as mean ± SEM in tables and figures. Analysis of variance was used to detect changes in a variable with time. Student’s t tests were used to compare measurements of the same parametric variable between subgroups. χ2 tests were used to compare categorical variables, e.g., the number of patients in ‘lived’ and ‘died’ subgroups. Coagulation variables were analyzed using logistic regression to assess association of coagulopathy severity and survival outcome. Univariate logistic regression was used to identify risk factor variables having significant association with survival outcome. Analyses were done using SAS software, version 9.1 SP3 (SAS Institute Inc., Cary, NC). p < 0.05 was considered significant.


During a 51 month period ending January 2003, 97 patients were resuscitated using our ICU protocol and received MT. Table 1 depicts the cohort as severely injured (Injury Severity Score [ISS], 29 ± 1); relatively young (age, 39 ± 2 years); mostly men (62%) who predominantly incurred blunt injury (73%); and demonstrating shock at hospital admission (ED BD, 10 ± 1 mEq/L). All patients required emergency OR and/or interventional radiology procedures and arrived at the ICU 6.8 ± 0.3 hours after ED admission. At start of ICU resuscitation, BD was 7 ± 1 mEq/L. Of note, coagulopathy was present at hospital admission (ED INR, 1.8 ± 0.2). In the ED, INR was obtained for 77 (79%) of these patients and, of these, 57 patients (74%) had INR > 1.2. At ICU admission, INR was obtained for all patients; INR was > 1.2 in 82 patients (85%). Comparison of ‘lived’ and ‘died’ subgroups showed that only severity of coagulopathy at ICU admission, indicated as INR, differed (p < 0.05; see Table 1). ED INR of ‘lived’ and ‘died’ subgroups were not significantly different. Pre-ICU resuscitation included 12 ± 1 units PRBC, 9 ± 1 L crystalloid fluid, and 5 ± 0.4 units FFP. This is consistent with the MT protocol by which FFP was not given until after 6 units PRBC, and the first blood bank response was a container with 4 units FFP and 6 units PRBC. Forty-nine patients (51%) received ≥10 units PRBC during ED, OR and/or interventional radiology procedures. Mean ICU length of stay (LOS) was 13 ± 1 days. Twenty-nine patients (30%) did not survive hospitalization. Six (6%) died during the 24-hour ICU resuscitation (three from exsanguination and three from fulminant early adult respiratory distress syndrome) and four (4%) died during ICU day 2 (two from exsanguination and two from early fulminant adult respiratory distress syndrome). Nineteen late deaths (20%) occurred due to multiple organ failure/sepsis (n = 11; ICU LOS 15 ± 3 days), adult respiratory distress syndrome (n = 4; ICU LOS 12 ± 5 days), withdrawal of support (n = 3; ICU LOS 20 ± 4 days), and pulmonary embolus (n = 1; ICU LOS, 22 days). Abdominal compartment syndrome occurred in seven of the patients who died. Overall, 5 of 29 deaths (17%) were due to hemorrhage during the first two hospital days. This subgroup with early death caused by hemorrhage did not differ from the overall group in ISS, ED INR, pre-ICU crystalloid fluid, PRBC or FFP volume, or ED to ICU admit time; however, they did have pre-ICU [Hb] less than [pre ICU [Hb] = 7.0 ± 1.1 v 10.3 ± 0.3 g/dL (p < 0.05)] and initial ICU INR greater than [initial ICU INR = 2.1 ± 0.2 versus 1.6 ± 0.04 (p < 0.05)] that of the overall group; indicating worsening coagulopathy during pre-ICU procedures for reasons that are unclear from clinical data.

Table 1
Table 1:
Description of Shock Resuscitation Massive Transfusion Cohort (p < 0.05)

Figure 1 summarizes ICU resuscitation interventions. All patients received LR and PRBC. Total volumes were 11 ± 1 L LR and 10 ± 1 units PRBC. Twenty-six patients (27%) received ≥10 units PRBC during ICU resuscitation. Figure 2 shows arterial pH and BD during ICU resuscitation. Severe acidosis was common at the start of resuscitation, with 7.00 ≤ pH ≤ 7.20 in 26 (27%) patients and pH < 7.00 in 3 patients (2%), and 8 ≤ BD ≤ 10 in 25 patients (26%) and BD > 10 in 25 patients (26%). Both mean pH and BD were within normal range within 8 hours. Figure 2 also shows T during ICU resuscitation. Of note, hypothermia was not a significant problem. At ICU admission, T was 35.4 ± 0.1 °C, with 32 ≤ T ≤ 35 °C in 27 patients (28%) and T < 32 °C in 1 patient (1%). Mean T increased rapidly to > 37 °C during the first 4 hours, and remained nearly constant for the duration of resuscitation.

Fig. 1.
Fig. 1.:
Cumulative volumes of packed red blood cells (PRBC) and lactated Ringer’s solution (LR) given during ICU day 1 as part of standardized shock resuscitation to attain and maintain O2 delivery index goal (600 mL/min-m2; 500 after January 2001).
Fig. 2.
Fig. 2.:
(A) Acidosis (arterial pH, base deficit [BD]) response to standardized ICU shock resuscitation, showing normalization within the first 8 hours in the ICU. (B) Core temperature (T) during resuscitation, indicating near normal temperature at ICU admission and rapid warming to T >37 °C during the first 4 hours in the ICU.

Coagulation data during ICU resuscitation are shown in Figures 3 (INR, PTT) and 4 ([plt], [fib]). At the start of ICU resuscitation, ∼7 hours after hospital admission, moderate coagulopathy persisted with INR (1.6 ± 0.1). At the start of ICU resuscitation, INR was >1.2 in 74 patients (76%) and INR was ≥1.4 in 55 patients (57%). Mean INR decreased to 1.4 ± 0.03 within 8 hours and remained nearly constant for the remaining 16 hours of ICU resuscitation. PTT was 58 ± 4 sec at the start and decreased to a stable plateau of 36 ± 1 sec within 12 hours during resuscitation. [plt] was 95 ± 9 kcells/mm3 at the start and, with the exception of a transient increase between 4 and 12 hours, tended to decrease during resuscitation. Severe thrombocytopenia was uncommon during the first 4 hours in the ICU, with 30 ≤ [plt] ≤ 50 kcells/mm3 in 12 patients (13%) and [plt] < 30 kcells/mm3 in 4 (4%) of these patients. [fib] was 143 ± 8 at the start and mean values increased in near linear fashion to 386 ± 17 mg/dL during resuscitation. During the first 4 hours in the ICU, [fib] less than normal limits (100 to 450 mg/dL) was uncommon, with 50 ≤ [fib] ≤ 80 mg/dL in 11 (11%) and [fib] < 50 mg/dL in 3 (3%) of these patients.

Fig. 3.
Fig. 3.:
(A) International normalized ratio (INR) during ICU day 1, indicating moderate coagulopathy (INR, 1.6 ± 0.1) at ICU admission and incomplete correction (INR, 1.4 ± 0.03) by the end of resuscitation and ICU day 1. (B) Partial thromboplastin time (PTT) during ICU day 1, indicating coagulopathy (PTT, 58 ± 4 sec) at ICU admission and incomplete correction (PTT, 36 ± 2 sec) by the end of resuscitation.
Fig. 4.
Fig. 4.:
(A) Platelet count ([plt]) during ICU day 1, showing adequate platelet concentration at ICU admission and maintenance >85 kcells/mm3, but tendency to decrease. (B) Fibrinogen concentration ([fib]) during ICU day 1, showing uniform, near-linear increase during ICU day 1.

Interventions for coagulopathy during ICU resuscitation are shown in Table 2. FFP was the most frequent intervention for coagulopathy, and 81 (84%) of these patients received 10 ± 1 units FFP. Fewer patients (57 [59%]) received plt component therapy. Cryoprecipitate was given to 20 (21%) of these patients.

Table 2
Table 2:
Interventions for Coagulopathy During Standardized ICU Shock Resuscitation

In this study cohort, univariate logistic regression analysis found severity of coagulopathy (indicated by INR) measured at arrival in the ICU to be associated with survival outcome (see Table 3; area under ROC, 0.71). Figure 5 depicts this relationship, with severe coagulopathy (INR ≥ 2.0) associated with ≥50% probability of death. An association was also found for severity of shock at ICU admission (indicated by BD) and mortality (area under ROC, 0.64); however, colinearity of INR and BD was apparent with multivariate analysis, indicating probable effect of acidosis on INR.

Table 3
Table 3:
Potential Risk Factors Tested for Association With Mortality Using Univariate Logistic Regression (p < 0.05)
Fig. 5.
Fig. 5.:
Univariate logistic regression analysis shows that severity of coagulopathy, indicated as international normalized ratio (INR), early after ICU admission is predictive of mortality (p = 0.02; area under ROC, 0.71).


It has long been recognized that the “bloody vicious cycle” of acidosis, hypothermia, and coagulopathy is an important factor in the early death of bleeding trauma patients who survive long enough to arrive at hospitals capable of providing life-saving hemorrhage control interventions.1,11 Despite tremendous advances in care at now-specialized Trauma Centers, hemorrhage remains the leading cause of early death.2,12 As we continue to study the epidemiology of patients who arrive with exsanguinating hemorrhage, it is apparent that a subset of these patients do not respond well to standard of care interventions. Because resuscitation is an obligatory intervention, we have focused our efforts on controlling ICU resuscitation by utilizing computerized decision support. With ongoing analyses of prospectively collected data, we have progressively refined this process of care and determined that the clinical trajectory of the nonresponder declares itself early in pre-ICU care.13 Therefore, we have developed several pre-ICU protocols to hasten identification and treatment of life-threatening hemorrhage and to optimize pre-ICU resuscitation. This analysis was undertaken to assess how well our empiric pre-ICU MT protocol is working in patients who are entered into our ICU resuscitation protocol.

This study cohort was severely injured patients with severe shock evident in the ED and persistent at ICU arrival. ICU resuscitation was effective. Mean CI increased to ∼4 L/min-m2 within 4 hours, [Hb] ∼ 11 g/dL remained stable, and BD was corrected within ∼8 hours. Although significant acidosis was present at ICU admission, it was reliably corrected with resuscitation.8,9,14 Together with BD, pH was normalized within 8 hours (see Fig. 2). To our surprise, severe hypothermia was uncommon at ICU admission. Only one patient arrived in the ICU with T < 32 °C. At the time of Shock Trauma ICU admission, T was 35.5 ± 0.1 °C and normalized within 4 hours despite continued need for large volume resuscitation and lack of active internal rewarming interventions (see Fig. 2).

These data indicate that, despite the absence of rigorous protocols in the ED and OR settings, acidosis and hypothermia were reasonably well managed at our Level I trauma center and did not seem to complicate our ICU resuscitation process. Coagulopathy, however, remained a significant problem. These patients arrived with coagulopathy (ED INR, 1.8 ± 0.2). Despite receiving quantities of blood products that were surprisingly consistent with our standard of care pre-ICU MT protocol (pre-ICU PRBC, 12 ± 1 units; pre-ICU FFP, 5 ± 0.4 units), ICU admission INR was 1.6 ± 0.04. Additionally, despite routine serial coagulation analyses and interventions directed by the bedside ICU physician, coagulopathy was not definitively corrected. INR decreased to a stable plateau of 1.4 ± 0.03 within 8 hours, and this moderate coagulopathy persisted for the remaining 16 hours of resuscitation (see Fig. 3). Transfusion of FFP was the primary intervention for coagulopathy correction (see Table 2). The ratio of units of FFP:PRBC during ICU resuscitation was 1:1, which exceeds published recommendations.7,15 The standard of care throughout the study period was to maintain [plt] ≥100 k cells/mm3 during active resuscitation. Forty-two patients (43%) had [plt] <100 k cells/mm3 during the first 4 hours in the ICU and, after a transient increase between 4 and 12 hours, mean [plt] tended to decrease as resuscitation proceeded (see Fig. 4). Mean [fib] increased steadily during ICU resuscitation with remarkable uniformity and remained within normal range of 100 to 450 mg/dL (see Fig. 4). Appropriately, few patients received cryoprecipitate, a more definitive intervention to increase [fib] than FFP (see Table 2). This challenges the recent emphasis for early cryoprecipitate administration. Although Fries et al.16 reported that increasing [fib] to greater than normal range decreases blood loss in an animal model of liver injury, clinical data are lacking.

We think that failure to correct coagulopathy during ICU resuscitation was largely attributable to inadequate pre-ICU intervention. The patients who arrived with obvious coagulopathy required ongoing transfusion and received FFP and PRBC according to our standard of care both pre-ICU (MT protocol, FFP withheld until after 6 units PRBC) and during ICU resuscitation (FFP:PRBC given 1:1). During ICU resuscitation, however, these patients had ongoing blood loss as indicated by ongoing PRBC transfusion requirements. Superposition of INR and PRBC transfusion data during ICU resuscitation, shown in Figure 6, clearly shows that significant coagulopathy was present and that these patients were receiving vigorous PRBC transfusion to maintain hemoglobin concentration. These observations suggest that more aggressive pre-ICU intervention to correct coagulopathy may be effective in decreasing PRBC requirement during ICU resuscitation and, because of the association with increased mortality, could improve outcome.17

Fig. 6.
Fig. 6.:
International normalized ratio (INR) and cumulative PRBC transfusion during ICU resuscitation showing persistent moderate coagulopathy (INR ∼1.4) and concurrent ongoing transfusion requirement.

Recent publications from other Level I trauma centers also identify coagulopathy as a significant problem in the early management of the trauma patient. Although traditionally attributed to hemodilution, acidosis, and hypothermia, two recent studies indicate that coagulopathy starts very soon after trauma, independent of these aggravating events. A retrospective trauma registry review from the Ryder Trauma Center (Miami, FL) documented 28% incidence of coagulopathy (defined as PT ≥14 sec) in trauma patients (median ISS, 9) at arrival to the trauma bay.5 A second report reviewed helicopter transports shortly after arrival to the Royal London Hospital (London, England, UK), and found 24% of trauma patients (median ISS, 20) to be coagulopathic (defined as PT > 18 sec in 16%, PTT > 60 sec in 13%, and thrombin time > 15 sec in 14%). These investigators also described a linear relationship between early coagulopathy, ISS, and mortality. Cosgriff et al.,18 from Denver Health Medical Center (Denver, CO), analyzed a transfusion registry designed to document coagulopathy as patients proceed from ED to OR and then to ICU. In these severely injured patients (mean ISS, 31 ± 2) who received >10 units PRBC in the first 24 hospital hours after injury, these researchers documented a 47% incidence of severe coagulopathy in the OR (defined as PT and PTT 2 times clinical laboratory normal). Using multiple logistic regression, these investigators identified four independent risk factors for severe coagulopathy (with odds ratios): (1) pH < 7.10 (12.3); (2) T < 34 °C (8.7); (3) ISS > 25 (7.7); and (4) systolic blood pressure < 70 mm Hg (5.8). Hirshberg et al.7 reported a computer simulation using data from exsanguinating patients treated at Ben Taub General Hospital (Houston, TX). The model accounted for blood component replacement, resuscitation-induced hemodilution, bleeding, and hemodynamic status. These authors concluded that existing protocols underestimate the dilution of clotting factors in severely bleeding patients. They recommended that FFP should be administered in a ratio of 2:3 with PRBC, or 2 units FFP should be administered concurrently with the first units of PRBC if severe hemorrhage is anticipated. This practice of early FFP administration is supported by another recent report from Denver Health Medical Center (Denver, CO), in which Biffl et al.19 described the ongoing refinement of a clinical pathway for management of hemodynamically unstable patients with pelvic fractures. The initial protocol emphasized advanced trauma life support with early hemorrhage control by a combination of external fixation and angio embolization. The next rendition, implemented 5 years later, emphasized earlier pelvic fracture stabilization by pelvic binding, implemented a MT protocol that started FFP transfusion in the ED, and minimized indiscriminant crystalloid fluid infusion. After implementation of the refined protocol, the authors reported a significant decrease in the incidence of death caused by exsanguination from 9% to 1%. Additionally, a recent consensus conference to address issues related to early MT after injury confirmed the need for earlier interventions in the massively injured patient who presents with shock, and concluded that the blood products currently available to emergently support the severely injured patient are inadequate and that focus needs to be on early coagulopathy correction.15,17,20–23

Based on our experience and the above cited studies, we think that coagulopathy remains a significant problem with severely injured patients and that it is present early after ED admission. The inciting mechanism for coagulopathy is not clear from the clinical data obtained, but the data do support the need for early recognition of coagulopathy and correction before ICU admission, and possibly for new concepts to address this issue.23

Therefore, we are developing and implementing a standardized protocol for prevention and correction of coagulopathy that starts in the ED. We have developed a multiple-tier protocol. The first interventions are empirically directed by a pre-ICU MT policy. Our revised protocol emphasizes early FFP administration in a ratio of 1 unit FFP to 1 unit PRBC, beginning with the first unit of PRBC transfusion, and is invoked by the trauma surgeon calling the blood bank as soon as severe bleeding is recognized and the need for MT is anticipated by the trauma surgeon. Our blood bank now maintains 5 units of fresh thawed plasma that are immediately available for MT patients. This MT protocol remains in effect until the patient arrives in the Shock Trauma ICU, after which the empiric MT protocol is stopped by the bedside ICU physician calling the blood bank, and coagulopathy correction is then based on an ICU protocol with clinical laboratory measurements and specified interventions for coagulation variables. We estimate the cost of the MT protocol to be incidental to our Trauma Center and to the individual patient because of the small percentage of patients admitted to a Trauma Center who require MT and because of the potentially life-saving early preemptive intervention. Development of the ICU process has proceeded using principles set forth by Morris and colleagues24 and is based on literature review, the data presented here, a limited set of standard clinical measurements able to be repeated, and our trauma team consensus discussions. This ICU coagulopathy prevention and correction decision support protocol uses thresholds and directs component therapy interventions to correct coagulopathy, indicated as INR, [plt], PTT, and [fib] as needed. ICU protocol development, begun in 2002, has undergone extensive revision and is now being integrated with the standardized 24-hour ICU shock resuscitation protocol. The ICU coagulopathy correction protocol directs Factor VIIa as a final intervention on a compassionate-use basis if coagulopathy and life-threatening bleeding persist despite traditional component interventions. Further clinical trial is needed for this intervention. Results of the recent European-South African trial were mixed, with advantage shown for blunt but not penetrating trauma victims,25and there is no US Food and Drug Administration indication for Factor VIIa for traumatic shock.26 Factor VIIa is available according to hospital-specific criteria and after consultation with an on-call hematologist.


This study indicates that coagulopathy is a problem that appears in severely injured patients at admission to the ED, and is not corrected despite early correction of acidosis and hypothermia. Additionally, coagulopathy is not corrected using current pre-ICU MT guideline therapy or by the bedside clinician during ICU shock resuscitation despite identification by serial clinical laboratory analyses and aggressive component replacement. Reasons for this are unclear from the data obtained, but may include the inability to correct coagulopathy during ongoing resuscitation with crystalloid fluid and PRBC, or that assessment by the bedside clinician is to accept moderate coagulopathy as an alternative to continued aggressive blood product administration. Development of computerized decision support for this process will allow us to decipher this issue.

Our data show that uncorrected coagulopathy at ICU admission is associated with ongoing transfusion requirements and that the severity of coagulopathy is directly associated with increased risk of death. For trauma patients presenting with exsanguinating hemorrhage, coagulopathy correction beginning with aggressive FFP administration pre-ICU may improve ICU resuscitation response and outcome.


1. Kashuk JL, Moore EE, Millikan JS, Moore JB. Major abdominal vascular trauma-a unified approach. J Trauma. 1982;22:672–679.
2. Sauaia A, Moore FA, Moore EE, et al. Early predictors of postinjury multiple organ failure. Arch Surg. 1994;129:39–45.
3. Ledgerwood AM, Lucas CE. A review of studies on the effects of hemorrhagic shock and resuscitation on the coagulation profile. J Trauma. 2003;54(Suppl):S68–S74.
4. Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy. J Trauma. 2003;54:1127–1130.
5. MacLeod JB, Lynn M, McKenney MG, et al. Early coagulopathy predicts mortality in trauma. J Trauma. 2003;55:39–44.
6. McKinley BA, Gonzalez EA, Balldin BC, et al. Revisiting the “Bloody Vicious Cycle”. Shock. 2004;21(Suppl 2):47.
7. Hirshberg A, Dugas M, Banez EI, et al. Minimizing dilutional coagulopathy in exsanguinating hemorrhage: a computer simulation. J Trauma. 2003;54:454–463.
8. Marr AB, Moore FA, Sailors RM, et al. “Starling curve” generation during shock resuscitation: Can it be done? Shock. 2004;21:300–305.
9. McKinley BA, Kozar RA, Cocanour CS, et al. Normal vs. supranormal O2 delivery goals in shock resuscitation: The response is the same. J Trauma. 2002;53:825–842.
10. McKinley BA, Sailors RM, Glorsky SL, et al. Computer directed resuscitation of major torso trauma. Shock. 2001;15(Suppl):46.
11. Mikhail J. The trauma triad of death: hypothermia, acidosis, and coagulopathy. AACN Clin Issues. 1999;10:85–94.
12. Sauaia A, Moore FA, Moore EE, et al. Multiple organ failure can be predicted as early as 12 hours postinjury. J Trauma. 1998;45:291.
13. Balogh Z, McKinley BA, Holcomb JA, et al. Both primary and secondary abdominal compartment syndrome can be predicted early and are harbingers of multiple organ failure. J Trauma. 2003;54:848–859.
14. McKinley BA, Marvin RG, Cocanour CS, et al. Blunt trauma resuscitation: the old can respond. Arch Surg. 2000;48:637–642.
15. Ketchum L, Hess JR, Hiippala S. Indications for early FFP, cryoprecipitate and platelet transfusions in trauma. J Trauma. 2006;60(Suppl 6):S51–S58.
16. Fries D, Krismer A, Klingler A, et al. Effect of fibrinogen on reversal of dilutional coagulopathy: a porcine model. Br J Anaesth. 2005;95:172–177.
17. Napolitano L. Cumulative risks of early RBC transfusion. J Trauma. 2006;60(Suppl 6):S26–34.
18. Cosgriff N, Moore EE, Sauaia A, et al. Predicting life-threatening coagulopathy in the massively transfused trauma patient: hypothermia and acidosis revisited. J Trauma. 1997;42:857–862.
19. Biffl WL, Smith WR, Moore EE, et al. Evolution of a multidisciplinary clinical pathway for the management of unstable patients with pelvic fractures. Ann Surg. 2001;233:843–850.
20. Holcomb JB, Hess JR (eds.). Early massive transfusion: state of the art—Conference Proceedings. US Army Institute of Surgical Research, Ft. Sam, Houston. TX. May 26–27, 2005 J Trauma. 2006;60(Suppl 6):S1–S2.
21. Holcomb JB, Hoyt D, Hess JR. Damage control resuscitation: the need for specific blood products to treat the coagulopathy of trauma. Transfusion. 2006;46:685–686.
22. Armand R, Hess JR. Treating coagulopathy in trauma patients. Transfus Med Rev. 2003;17:223–231.
23. Malone DL, Hess JR, Fingerhut A. Massive transfusion practices around the globe and a suggestion for a common massive transfusion protocol. J Trauma. 2006;60(Suppl 6):S91–S96.
24. Morris AH. Algorithm-based decision making. In: Tobin MJ, ed. Principles and Practice of Intensive Care Monitoring. New York: McGraw-Hill; 1997:1355–1381.
25. Boffard KD, Warren B, Iau P. Decreased transfusion utilization and improved outcome associated with the use of recombinant Factor VIIa as an adjunct in trauma. J Trauma. 2004;57:451.
26. Enomoto TM, Thorborg P. Emerging off-label uses for recombinant activated factor VII: grading the evidence. Crit Care Clin. 2005;21:611–632.
© 2007 Lippincott Williams & Wilkins, Inc.