Initial Resuscitation and Prevention of Further Bleeding
Minimal Elapsed Time
Recommendation 1
We recommend that severely injured patients be transported directly to an appropriate trauma facility. (Grade 1B)
We recommend that the time elapsed between injury and bleeding control be minimized. (Grade 1A)
Rationale
In managing severe trauma, time is of the essence. The organization of trauma care systems is predicated on the understanding that prompt, specialized multidisciplinary care significantly improves patient outcomes. Trauma networks are designed to facilitate the swift transfer of severely injured individuals to centers equipped to handle complex injuries, ensuring immediate access to critical interventions. This approach, while seemingly straightforward, involves navigating the inherent tension between rapid transport and the immediate need for life-saving procedures at the scene. The heterogeneity in trauma care systems across and even within countries highlights the ongoing debate and diverse strategies employed to optimize patient flow and resource allocation in trauma management.
While robust, high-level evidence remains somewhat limited due to the inherent complexities of trauma research, a broad consensus supports the effectiveness of organized trauma systems. Studies suggest that the establishment of trauma systems is associated with an approximate 15% reduction in overall trauma mortality and a remarkable 50% decrease in preventable deaths [37–39]. This underscores the critical role of system-level organization in enhancing survival rates for trauma patients.
Interestingly, inter-hospital transfers, while potentially adding to the time elapsed before definitive care, do not appear to negatively impact overall mortality [40]. This may suggest that the benefits of specialized trauma center care outweigh the potential risks associated with transfer. However, the literature neither definitively supports nor refutes the concept of direct transport from the accident scene to a major trauma center compared to initial stabilization at a secondary center [41]. This area warrants further investigation to refine pre-hospital protocols and optimize triage decisions.
Emerging evidence suggests that age plays a crucial role in determining the threshold for trauma center care. For patients older than 65 years, a lower threshold for trauma center referral may be beneficial [42]. This reflects the physiological vulnerabilities of older adults and their potentially reduced resilience to traumatic injuries. Conversely, the relationship between a hospital’s trauma patient volume and outcomes remains inconclusive, with no definitive link established between higher volume and improved survival [43]. This suggests that factors beyond volume, such as system organization, resource availability, and expertise, may be more critical determinants of patient outcomes.
Despite the ongoing need for more definitive evidence, a fundamental principle of trauma care remains unchallenged: “systemized” trauma care, which strategically matches patients to the most appropriate treatment facility, is advantageous. The definition of “appropriate” is nuanced and multifactorial, depending on the patient’s specific profile, the nature and severity of their injuries, and the capabilities of available hospital facilities. This individualized approach to trauma triage and transport is paramount in optimizing resource utilization and ensuring timely access to the right level of care.
For trauma patients experiencing ongoing hemorrhage requiring emergency surgery, minimizing the time from injury to operative intervention is directly linked to increased survival. Alarmingly, over half of trauma fatalities occur within the first 24 hours post-injury [6], emphasizing the critical importance of rapid hemorrhage control. While prospective randomized controlled trials (RCTs) are ethically and logistically challenging in this context, well-designed retrospective studies consistently demonstrate the survival benefit of early surgical intervention in traumatic hemorrhagic shock [44–46]. Furthermore, indirect evidence from trauma system analyses reinforces the significance of minimizing the time interval between hospital admission and surgical bleeding control in these critical patients [47, 48].
The principle of minimizing time to surgery in trauma care is a cornerstone of best practice, so firmly established that it is unlikely to be subjected to clinical trials due to the absence of equipoise. The ethical imperative to provide the most effective and timely care in trauma hemorrhage dictates the continued adherence to this critical recommendation.
Tourniquet Use
Recommendation 2
We recommend adjunct tourniquet use to stop life-threatening bleeding from open extremity injuries in the pre-surgical setting. (Grade 1B)
Rationale
In the context of uncontrolled arterial bleeding from severe extremity injuries, such as those resulting from penetrating trauma, blast injuries, or traumatic amputations, tourniquet application emerges as a simple yet highly effective method for immediate hemorrhage control [49–53]. Tourniquets have become a standard of care, particularly in military settings, for managing severe external hemorrhage following combat injuries. Numerous publications have documented the effectiveness of tourniquets in these specific scenarios, both in adult and pediatric populations [49–52, 54, 55].
Studies involving volunteers have demonstrated the efficacy of currently available tourniquet devices in achieving rapid hemorrhage control [53]. Importantly, these studies also highlight the ineffectiveness of “pressure point control” due to the rapid development of collateral circulation within seconds. Interestingly, tourniquet-induced pain was not a frequently reported issue among patients in these studies. Crucially, there is no evidence or rationale to support the use of tourniquets in closed injuries, where the source of bleeding is not externally accessible.
Once applied, tourniquets should remain in place until definitive surgical control of bleeding is achieved [50, 52]. However, the duration of tourniquet application should be minimized to mitigate potential complications. Prolonged or improper tourniquet placement can lead to nerve paralysis and limb ischemia [56], although these effects are relatively rare [54]. Some literature suggests a maximum application time of 2 hours [56], while reports from military experiences describe successful extremity salvage even with tourniquet times up to 6 hours [50].
The translation of military tourniquet experience to civilian practice has been a subject of considerable discussion, as civilian trauma often presents with different injury patterns and contexts. While direct pressure is sufficient to control bleeding in many civilian wounds, tourniquet use is strongly recommended for uncontrolled external bleeding from limb injuries, whether blunt [57] or penetrating [58] in nature. This recommendation underscores the importance of considering tourniquets as a vital adjunct in pre-hospital and early hospital management of life-threatening extremity hemorrhage in civilian trauma as well.
Ventilation
Recommendation 3
We recommend the avoidance of hypoxemia. (Grade 1A)
We recommend normoventilation of trauma patients. (Grade 1B)
We suggest hyperventilation in the presence of signs of imminent cerebral herniation. (Grade 2C)
Rationale
Tracheal intubation in severely injured patients is a critical intervention that demands careful consideration of both benefits and risks, requiring skilled operators and appropriate training. The primary goals of intubation are to secure a patent airway, ensure adequate ventilation, and optimize oxygenation. While certain situations unequivocally necessitate intubation, such as airway obstruction, altered consciousness (Glasgow Coma Scale (GCS) ≤8), hemorrhagic shock, hypoventilation, or hypoxemia [59], other factors must also be carefully weighed.
For instance, positive pressure ventilation, while crucial for respiratory support, can paradoxically induce life-threatening hypotension in hypovolemic patients [60]. Some studies have even reported increased mortality associated with pre-hospital intubation [61], highlighting the potential for adverse consequences if intubation is not performed judiciously and expertly.
Several factors influence the success of intubation and, consequently, patient prognosis. Rapid sequence induction (RSI) is generally considered the optimal method for securing the airway in trauma patients [62]. However, ongoing debates persist regarding key aspects of intubation management, including the optimal decision-making process for intubation, the selection of induction and maintenance drugs, the choice of rescue devices in case of failed intubation, and the ideal infrastructure for emergency airway management services. The majority of available data are derived from retrospective studies, inherently susceptible to bias, contributing to the continued controversy surrounding the appropriate utilization of tracheal intubation following traumatic injury [63].
The detrimental effects of hypoxemia, particularly in patients with traumatic brain injury (TBI), are well-established [64, 65]. Therefore, high concentrations of oxygen are typically administered to ensure adequate oxygen delivery to ischemic brain tissue during the initial management of TBI patients. However, some studies have suggested that excessive hyperoxia may paradoxically be associated with increased mortality [66]. The underlying mechanisms for this potential harm are not fully elucidated but may involve increased production of harmful free radicals or the exacerbation of hyperoxic vasoconstriction. Therefore, a cautious approach to oxygen administration, avoiding extreme hyperoxia, may be prudent. While the precise level of hyperoxia that becomes detrimental in trauma patients remains undefined, most studies consider a PaO2 exceeding 200–300 mmHg (27–40 kPa) to be potentially harmful [67, 68].
Adequate ventilation plays a crucial role in influencing outcomes for severely injured trauma patients. A common tendency among rescue personnel is to inadvertently hyperventilate patients during initial resuscitation [69, 70]. However, evidence suggests that hyperventilated trauma patients may experience increased mortality compared to those ventilated at normal rates [66]. The target PaCO2 range for trauma patients should ideally be 5.0–5.5 kPa (35–40 mmHg).
The effects of hyperventilation on bleeding and outcomes specifically in severe trauma patients without TBI remain less clear. Several potential mechanisms may mediate the adverse effects of hyperventilation and hypocapnia, including increased vasoconstriction leading to decreased cerebral blood flow and impaired tissue perfusion. Cerebral tissue lactic acidosis has been observed almost immediately following the induction of hypocapnia in both children and adults with TBI and hemorrhagic shock [71]. Even modest levels of hypocapnia can have detrimental effects [72]. Furthermore, in the setting of absolute or relative hypovolemia, excessive positive-pressure ventilation rates can further compromise venous return, potentially leading to hypotension and even cardiovascular collapse [73, 74].
The sole clinical scenario where hyperventilation-induced hypocapnia may have a limited role is in the presence of imminent cerebral herniation. The reduction in cerebral blood flow caused by acute hypocapnia during hyperventilation can transiently decrease intracranial pressure, potentially buying valuable time until other definitive measures can be implemented [75, 76]. Clinical signs such as unilateral or bilateral pupillary dilation or decerebrate posturing indicate an extreme risk of impending death or irreversible brain damage. In such critical situations, hyperventilation may be employed as a temporizing measure to gain time for more definitive interventions. While clinical studies specifically evaluating this practice are lacking, the physiological rationale is sound, and the risk-benefit balance appears favorable given the dire prognosis without intervention. However, it is crucial to normalize PaCO2 as soon as clinically feasible once the immediate threat of herniation has been addressed.
Ventilation strategies employing low tidal volumes (6 ml/kg) are recommended for patients with or at risk of acute respiratory distress syndrome (ARDS) [77]. In patients with normal lung function, the optimal tidal volume remains more debated, but growing evidence suggests that the injurious effects of high tidal volumes may manifest very early in the course of ventilation. Randomized studies have demonstrated that short-term ventilation with lower tidal volumes is associated with reduced systemic markers of inflammation and improved lung mechanics [78]. Although further research is warranted, the early adoption of protective ventilation strategies, incorporating low tidal volumes and moderate positive end-expiratory pressure (PEEP), is generally recommended, especially in bleeding trauma patients who are inherently at increased risk of developing ARDS.
Diagnosis and Monitoring of Bleeding
Initial Assessment
Recommendation 4
We recommend that the physician clinically assess the extent of traumatic hemorrhage using a combination of patient physiology, anatomical injury pattern, mechanism of injury and the patient’s response to initial resuscitation. (Grade 1C)
Rationale
While overt blood loss may be readily apparent in some trauma cases, relying solely on visual estimation or physiological parameters to gauge the degree of bleeding can be unreliable [79]. The mechanism of injury serves as a vital screening tool for identifying patients at higher risk of significant hemorrhage. For example, the American College of Surgeons defines a fall from a height of 6 meters (20 feet) or greater as a “critical falling height,” indicative of a mechanism associated with major injuries [80]. Other critical mechanisms include high-energy deceleration impacts, distinctions between low-velocity versus high-velocity gunshot wounds, and crush injuries.
The mechanism of injury, in conjunction with the overall injury severity, the patient’s physiological presentation, and their response to initial resuscitation efforts, should guide the decision to initiate early surgical bleeding control, as outlined in the Advanced Trauma Life Support (ATLS) protocol [81–84]. Table 2 of the original article provides a summary of estimated blood loss based on initial patient presentation, categorized according to the ATLS classification system. The ATLS classification has been validated as a useful guide for quantifying blood loss with reasonable accuracy in hemorrhagic shock [85].
However, it is important to acknowledge the limitations of the ATLS classification. Several studies have highlighted discrepancies in the weighting assigned to individual parameters within the ATLS system when assessing blood loss, making consistent patient classification challenging. Mutschler et al. found that over 90% of trauma patients could not be definitively categorized using the ATLS classification of hypovolemic shock [86]. Furthermore, the same group concluded that the ATLS system may underestimate the degree of mental status impairment associated with hypovolemic shock and overestimate the degree of tachycardia associated with hypotension [87]. A retrospective analysis of the ATLS classification’s validity demonstrated that while increasing blood loss does correlate with increases in heart rate and decreases in blood pressure, the magnitude of these changes is often less pronounced than suggested by the ATLS classification. Additionally, significant changes in respiratory rate or level of consciousness may not consistently correlate with the degree of bleeding [88].
Table 3 of the original article characterizes the three distinct patterns of response to initial fluid resuscitation: rapid responders, transient responders, and non-responders. Transient responders and non-responders are considered high-risk candidates for immediate surgical bleeding control due to their failure to achieve or sustain hemodynamic stability with initial fluid resuscitation alone.
Specific scoring systems designed to predict the risk of hemorrhagic shock can be valuable tools in facilitating prompt and appropriate treatment decisions. The shock index (heart rate divided by systolic blood pressure) has demonstrated utility in predicting critical bleeding risk [89] and can aid in identifying trauma patients who will likely require hemostatic interventions [90]. Paladino et al. [91] found that while the shock index can effectively highlight abnormal values, it is insufficiently sensitive to definitively rule out major injury and should not diminish clinical suspicion in high-risk scenarios.
The Trauma-Associated Severe Hemorrhage (TASH) score utilizes seven readily available parameters [systolic blood pressure, hemoglobin (Hb), presence of intra-abdominal fluid, complex long bone and/or pelvic fractures, heart rate, base excess, and gender] to predict the probability of requiring massive transfusion. Maegele et al. [92] validated the TASH score in a retrospective analysis of a large dataset from the German Trauma Registry, confirming its ability to predict individual probability of massive transfusion and, by extension, ongoing life-threatening hemorrhage. The TASH score was subsequently re-validated using an even larger cohort of over 5,800 patients from the same registry [93], further solidifying its role as a valuable predictive tool in trauma hemorrhage management.
Immediate Intervention
Recommendation 5
We recommend that patients presenting with hemorrhagic shock and an identified source of bleeding undergo an immediate bleeding control procedure unless initial resuscitation measures are successful. (Grade 1B)
Rationale
In many cases, the source of traumatic bleeding is immediately obvious, particularly in penetrating injuries, which are more likely to necessitate surgical bleeding control. A retrospective study analyzing 106 abdominal vascular injuries revealed that all 41 patients arriving in shock following gunshot wounds were appropriately triaged for rapid transfer to the operating room for surgical hemorrhage control [94]. Similar findings from a study of 271 patients undergoing immediate laparotomy for gunshot wounds reinforce the need for early surgical intervention in patients with these injury patterns and signs of severe hypovolemic shock. This principle extends, albeit to a lesser degree, to abdominal stab wounds [95]. Data from injuries sustained from penetrating metal fragments during the Vietnam War further validate the critical need for early surgical control when patients present in shock [96].
In blunt trauma, the mechanism of injury can provide valuable insights into the likelihood of requiring surgical bleeding control in patients presenting with hemorrhagic shock. While studies specifically examining the relationship between mechanism of injury and bleeding risk in blunt trauma are limited, and high-level evidence from randomized prospective trials is lacking [97], certain injury patterns are clearly associated with a higher propensity for significant hemorrhage. Objective data describing the precise relationship between mechanism of injury and bleeding risk in skeletal fractures, particularly long bone fractures, are currently limited.
Traffic accidents are the leading cause of pelvic injuries, accounting for approximately 60% of pelvic fractures, followed by falls from height (23%). The remaining cases are primarily attributed to motorbike collisions and vehicle-pedestrian accidents [98, 99]. A well-established correlation exists between “unstable” pelvic fractures and associated intra-abdominal injuries [98, 100]. Major pelvic fractures are also frequently associated with severe head injuries, concomitant thoracic, abdominal, urological, and other skeletal injuries [98]. High-energy injuries, such as those sustained in motor vehicle crashes at significant speed, tend to cause more extensive damage to both the pelvis and surrounding organs. Patients with high-energy pelvic injuries often require larger volumes of blood transfusion, and over 75% may have concurrent injuries to the head, thorax, abdomen, or genitourinary system [101].
It is well-documented that “unstable” pelvic fractures are strongly associated with massive hemorrhage [100, 102], with hemorrhage being the leading cause of mortality in patients with major pelvic fractures. Vertical shear pelvic ring fractures, particularly those with caudal displacement of the hemipelvis, can cause significantly greater disruption to the pelvic floor and pelvic vasculature compared to standard vertical shear injuries. Inferior displacement of the hemipelvis observed on X-ray imaging should raise immediate suspicion for severe arterial injuries [103].
In blunt chest trauma, hemothoraces exceeding 500 ml typically warrant chest tube insertion. Thoracotomy is indicated for ongoing bleeding, defined as chest tube output greater than 1500 ml within 24 hours or greater than 200 ml for 3 consecutive hours. Acute damage control thoracotomy should be considered in cases of refractory hemorrhagic shock due to persistent chest bleeding, especially when initial chest tube output is greater than 1500 ml [104, 105].
Further Investigation
Recommendation 6
We recommend that patients presenting with hemorrhagic shock and an unidentified source of bleeding undergo immediate further investigation. (Grade 1B)
Rationale
Patients presenting with hemorrhagic shock and an unidentified source of bleeding require immediate and systematic investigation to locate the source of hemorrhage and guide targeted interventions. The chest, abdominal cavity, and pelvic ring represent the major potential spaces for acute blood loss in trauma. Beyond thorough clinical examination, initial diagnostic modalities during the primary survey should include chest and pelvic X-rays, in conjunction with focused ultrasonography [106], as recommended by ATLS guidelines [84, 107, 108].
In specialized trauma centers with readily available computed tomography (CT) scanners [109], CT imaging may supplant conventional radiographic techniques during the primary survey. Huber-Wagner et al. demonstrated in a multicenter study involving over 8,000 adult major blunt trauma patients that close proximity of the CT scanner to the trauma room significantly improved survival rates [110]. The authors recommend that emergency department planning prioritize the placement of CT scanners either directly within the trauma room or within 50 meters to facilitate rapid diagnostic imaging.
Conversely, a systematic literature review by Jorgensen and colleagues found no conclusive evidence that pre-hospital ultrasound of the abdomen or chest improves the treatment of trauma patients [111]. This highlights the ongoing debate and need for further research to define the optimal role of pre-hospital ultrasound in trauma assessment and triage.
Imaging
Recommendation 7
We recommend early imaging (ultrasonography or contrast-enhanced CT) for the detection of free fluid in patients with suspected torso trauma. (Grade 1B)
Intervention
Recommendation 8
We recommend that patients with significant intra-thoracic, intra-abdominal or retroperitoneal bleeding and hemodynamic instability undergo urgent intervention. (Grade 1A)
Further Assessment
Recommendation 9
We recommend CT assessment for hemodynamically stable patients. (Grade 1B)
Rationale
Blunt abdominal trauma poses a significant diagnostic challenge and is a frequent source of occult internal bleeding. Ultrasonography has become a well-established, rapid, and non-invasive diagnostic tool for detecting intra-abdominal free fluid in the emergency department [112–114]. Large prospective observational studies have consistently demonstrated high specificity and accuracy of initial ultrasound examination for detecting intra-abdominal injuries in both adult and pediatric trauma populations, although sensitivity, particularly for solid organ injury, can be lower [115–121]. Liu and colleagues [122] reported high sensitivity, specificity, and accuracy of initial ultrasound examination specifically for the detection of hemoperitoneum.
While ultrasonography exhibits high specificity, its sensitivity for detecting free intraperitoneal fluid in penetrating torso trauma [123] and blunt abdominal trauma in children [124] can be lower. A positive ultrasound finding is highly suggestive of hemoperitoneum, but a negative initial abdominal ultrasound should prompt further diagnostic investigations, particularly if clinical suspicion remains high.
The role of CT scanning in the acute trauma setting is extensively documented [125–132], with multislice computed tomography (MSCT) becoming increasingly integral to modern trauma imaging protocols. The integration of advanced MSCT scanners within or immediately adjacent to the emergency room allows for rapid assessment of trauma patients upon arrival [127, 128]. Modern MSCT scanners can achieve total whole-body scanning times of less than 30 seconds, enabling swift and comprehensive diagnostic evaluation. Weninger and colleagues [128] demonstrated in a retrospective study comparing 370 patients that faster diagnosis using MSCT led to reduced emergency room time, operating room time, and intensive care unit (ICU) length of stay [128]. Huber-Wagner et al. [109] also highlighted the survival benefit associated with integrating whole-body CT into early trauma care. CT diagnosis significantly increases the probability of survival in polytrauma patients [110]. Whole-body CT as a standard diagnostic modality during the earliest resuscitation phase for polytrauma patients offers the added advantage of simultaneously identifying head and chest injuries, as well as other potential bleeding sources in multiply injured individuals.
Contrast medium-enhanced CT scanning offers further diagnostic refinement. Anderson et al. [133, 134] demonstrated high accuracy in evaluating splenic injuries resulting from trauma using intravenous (i.v.) contrast material. Delayed-phase CT imaging can be particularly useful in detecting active bleeding from solid organs. Fang et al. [135] showed that pooling of contrast material within the peritoneal cavity in blunt liver injuries strongly indicates active and massive bleeding, often requiring emergent surgical intervention. Intraparenchymal pooling of contrast material with an intact liver capsule, however, often suggests self-limited hemorrhage, and these patients may respond well to non-operative management. Tan and colleagues [136] found that preoperative CT scans in patients with hollow viscus and mesenteric injuries following blunt abdominal trauma typically exhibit abnormalities. Wu et al. [137] demonstrated the accuracy of CT in identifying severe, life-threatening mesenteric hemorrhage and blunt bowel injuries.
Compared to MSCT, traditional diagnostic and imaging techniques have inherent limitations. The diagnostic accuracy, safety, and effectiveness of immediate MSCT are contingent upon sophisticated pre-hospital treatment by experienced emergency medical personnel and rapid transportation times [138, 139]. If MSCT is not readily available within the emergency room, CT scanning necessitates patient transport to a dedicated CT room, requiring careful consideration of the potential risks and benefits of the procedure. During transport, continuous monitoring of vital signs and ongoing resuscitation efforts are paramount. For patients with questionable hemodynamic stability, imaging techniques such as ultrasound and chest and pelvic radiography may be more appropriate initial diagnostic modalities. Peritoneal lavage is rarely indicated when ultrasound or CT imaging capabilities are available [140]. The time required for transfer to and from all forms of diagnostic imaging must be carefully considered, particularly in hemodynamically unstable patients. In addition to initial clinical assessment, point-of-care testing results, including complete blood count, hematocrit (Hct), blood gases, and lactate levels, should ideally be readily accessible to inform rapid clinical decision-making.
Hypotensive patients (systolic blood pressure below 90 mmHg) with evidence of free intra-abdominal fluid on ultrasonography or CT are potential candidates for early surgical intervention if they cannot be stabilized with initial fluid resuscitation [141–143]. A retrospective study by Rozycki and colleagues [144] involving over 1,500 patients (both blunt and penetrating trauma) assessed with ultrasound as an early diagnostic tool showed that ultrasound examination exhibited sensitivity and specificity approaching 100% in hypotensive patients.
Many patients presenting with free intra-abdominal fluid on ultrasound can safely undergo further evaluation with MSCT. Generally, adult patients should be hemodynamically stable before undergoing MSCT outside of the emergency room setting [144]. Hemodynamically stable patients with high-risk mechanisms of injury, such as high-energy trauma, or even low-energy injuries in elderly individuals, should be considered for MSCT after ultrasound to comprehensively assess for additional injuries. With increasing integration of CT scanners within resuscitation units, whole-body CT diagnosis may eventually supplant ultrasound as the primary diagnostic modality in certain trauma centers.
MSCT is considered the gold standard for identifying retroperitoneal hemorrhage (RPH). Contrast-enhanced CT can detect RPH in 100% of cases and may identify the source of bleeding in approximately 40% of cases through visualization of contrast extravasation [145].
Hemodynamically unstable patients with significant intrathoracic, intra-abdominal, or retroperitoneal bleeding may require urgent intervention. In thoracic trauma with chest bleeding, chest tube insertion is typically the initial surgical step, often preceding acute damage control thoracotomy. Surgical bleeding control is generally necessary in unstable patients with hemoperitoneum. Patients with pelvic trauma and significant retroperitoneal hematoma may require external pelvic compression, retroperitoneal packing, or urgent radiologic embolization to achieve pelvic hemorrhage control [146–148].
Hemoglobin
Recommendation 10
We recommend that a low initial Hb be considered an indicator for severe bleeding associated with coagulopathy. (Grade 1B)
We recommend the use of repeated Hb measurements as a laboratory marker for bleeding, as an initial Hb value in the normal range may mask bleeding. (Grade 1B)
Rationale
Hemoglobin (Hb) or hematocrit (Hct) assays are fundamental components of the basic diagnostic workup for trauma patients. While Hct, a calculated parameter derived from Hb, was frequently analyzed in earlier studies, Hb is now more widely used in clinical practice. For consistency and clarity, these guidelines will refer to both parameters interchangeably, aligning with the specific parameter reported in the cited literature.
The diagnostic value of Hb or Hct in identifying trauma patients with severe injury and occult bleeding sources has been a subject of ongoing debate [149–151]. A significant limitation of Hb/Hct’s diagnostic utility is the confounding effect of resuscitation measures, particularly intravenous fluid administration and erythrocyte concentrate transfusion [152–154]. Furthermore, initial Hb or Hct values may not accurately reflect the extent of blood loss immediately after injury because patients initially lose whole blood, and compensatory mechanisms involving fluid shifts from the interstitial space take time to manifest and may not be reflected in early measurements.
The concept of low sensitivity of initial Hb/Hct in detecting severe bleeding has been challenged by some studies. Ryan et al. [155] found in a retrospective study of 196 trauma patients that Hct at admission demonstrated a strong correlation with hemorrhagic shock. Other researchers have similarly advocated for a greater emphasis on initial Hct values in assessing blood loss in trauma patients. Thorson et al., in a retrospective analysis of nearly 1,500 consecutive trauma patients, reported that initial Hct was more strongly associated with the need for transfusion than other commonly used parameters, such as heart rate, blood pressure, or acidemia, suggesting that fluid shifts occur rapidly after trauma and highlighting the potentially more important role of Hct in initial trauma assessment [156]. An initially low Hb level is also a recognized predictive criterion for massive transfusion in scoring systems like the TASH [92] and Vandromme [157] scores.
Thorson et al. [158] analyzed changes in Hct in serial determinations and concluded that the change in Hct is a reliable parameter for detecting ongoing blood loss. Two prospective observational diagnostic studies also demonstrated the sensitivity of serial Hct measurements in identifying patients with severe injuries [149, 150]. Decreasing serial Hct measurements may indicate continued bleeding; however, patients with significant bleeding may maintain stable serial Hct values in the context of ongoing resuscitation and compensatory physiological mechanisms.
Acute anemia itself may negatively impact the clotting process. A low Hct can reduce platelet margination, potentially hindering platelet activation. Moreover, Schlimp et al. [159] demonstrated that fibrinogen levels below 150 mg/dl are detected in a substantial proportion (up to 73%) of patients with admission Hb levels lower than 10 g/dl, highlighting the frequent co-occurrence of anemia and hypofibrinogenemia in trauma-induced coagulopathy.
Serum Lactate and Base Deficit
Recommendation 11
We recommend serum lactate and/or base deficit measurements as sensitive tests to estimate and monitor the extent of bleeding and shock. (Grade 1B)
Rationale
Serum lactate has been utilized as a diagnostic parameter and prognostic marker of hemorrhagic shock since the 1960s [160]. Lactate production, a byproduct of anaerobic glycolysis, serves as an indirect indicator of oxygen debt, tissue hypoperfusion, and the severity of hemorrhagic shock [161–164]. Similarly, base deficit values, derived from arterial blood gas analysis, provide an indirect assessment of global tissue acidosis resulting from impaired perfusion [161, 163].
Vincent and colleagues [165] demonstrated the prognostic value of serial lactate measurements in predicting survival in patients with circulatory shock. Their prospective study showed that changes in lactate concentration provide an early and objective assessment of a patient’s response to therapeutic interventions and suggested that repeated lactate determinations are a reliable prognostic index for patients with circulatory shock [165]. Abramson and colleagues [166] further validated these findings in a prospective observational study of patients with multiple traumatic injuries, evaluating the correlation between lactate clearance and survival. All patients whose lactate levels normalized (≤2 mmol/l) within 24 hours survived. Survival rates decreased to 77.8% if normalization occurred within 48 hours and dramatically to 13.6% in patients with persistently elevated lactate levels (>2 mmol/l) beyond 48 hours [166]. Manikis et al. [167] corroborated these findings, showing that initial lactate levels were higher in non-survivors of major trauma and that prolonged time to lactate normalization (>24 hours) was associated with the development of post-traumatic organ failure [167].
Lactate and/or base deficit measurements may be particularly valuable in penetrating trauma. In this type of injury, traditional triage vital signs, such as blood pressure, heart rate, and respiratory rate, may not accurately reflect the severity of injury and often do not correlate well with lactate or base deficit levels [168].
The reliability of lactate determination may be compromised when traumatic injury is associated with alcohol consumption. Ethanol metabolism leads to the conversion of pyruvate to lactate via lactate dehydrogenase, artificially elevating blood lactate levels. In alcohol-associated trauma, base deficit may be a more reliable prognostic indicator than lactate [169], although some authors suggest that ethanol-induced acidosis may also affect base deficit, potentially masking the true prognostic significance in trauma patients [170]. Therefore, lactate measurements in the context of alcohol-related trauma should be interpreted with caution.
Similar to lactate levels, initial base deficit, obtainable from either arterial or peripheral venous blood [171], has emerged as a potent independent predictor of mortality in patients with traumatic hemorrhagic shock [169]. Davis and colleagues [172] categorized base deficit severity into mild (-3 to -5 mEq/l), moderate (-6 to -9 mEq/l), and severe (>−10 mEq/l) categories, demonstrating a clear correlation between increasing base deficit severity and escalating mortality rates [172]. The same research group further showed that base deficit is a superior prognostic marker of death compared to pH in arterial blood gas analyses [173]. Mutschler et al. [174], analyzing a large cohort of over 16,000 severely injured patients from the German Trauma Registry, concluded that base deficit determination upon emergency department admission predicts transfusion requirements and mortality more effectively than the ATLS classification [174]. Furthermore, base deficit has been shown to be a highly sensitive marker for the extent of post-traumatic shock and mortality in both adult and pediatric trauma populations [175, 176].
Despite the robust data supporting the prognostic value of lactate levels in hemorrhagic shock, large-scale prospective studies specifically examining the correlation between base deficit and outcome are still needed. While both base deficit and serum lactate levels are well-correlated with shock and resuscitation status, they do not always strictly correlate with each other in severely injured patients [177]. Therefore, independent assessment of both parameters is recommended for a comprehensive evaluation of shock severity in trauma patients [161, 163, 177].
Coagulation Monitoring
Recommendation 12
We recommend that routine practice include the early and repeated monitoring of coagulation, using either a traditional laboratory determination [prothrombin time (PT), activated partial thromboplastin time (APTT) platelet counts and fibrinogen] (Grade 1A) and/or a viscoelastic method. (Grade 1C)
Rationale
Standard coagulation monitoring in trauma patients includes early and repeated measurements of prothrombin time (PT), activated partial thromboplastin time (APTT), platelet counts, and fibrinogen levels. Increasing emphasis is placed on the critical roles of fibrinogen and platelet measurements in the context of trauma-induced coagulopathy. While conventional coagulation screens, such as the international normalized ratio (INR) and APTT, are routinely used to “monitor” coagulation, it is crucial to recognize that these tests primarily assess only the initiation phase of blood coagulation, representing a mere fraction (approximately 4%) of total thrombin generation [178]. Consequently, conventional coagulation screens may appear deceptively normal even when the overall hemostatic capacity is significantly compromised [13, 179–183].
Furthermore, delays in detecting traumatic coagulopathy can negatively impact patient outcomes. Viscoelastic testing methods, such as thromboelastography (TEG) and rotational thromboelastometry (ROTEM), offer a significantly shorter turnaround time compared to conventional laboratory testing, potentially saving 30–60 minutes in critical diagnostic delays [181, 184, 185]. Viscoelastic testing may also be valuable in detecting coagulation abnormalities associated with the use of direct thrombin inhibitors, such as dabigatran, argatroban, bivalirudin, or hirudin. Moreover, early parameters of clot firmness assessed by viscoelastic testing have demonstrated predictive value for the need for massive transfusion, the incidence of thrombotic/thromboembolic events, and mortality in surgical and trauma patients [181, 186–195]. Therefore, comprehensive and rapid monitoring of blood coagulation and fibrinolysis using viscoelastic methods may facilitate more targeted and effective hemostatic therapy compared to relying solely on conventional laboratory tests.
Point-of-care tools, including thromboelastometry and portable coagulometers, have been developed to enable rapid coagulopathy detection in the emergency room or at the patient bedside, providing real-time data to guide patient management. Portable coagulometers providing INR or APTT results have demonstrated acceptable accuracy for point-of-care INR testing in the emergency department compared to traditional laboratory methods [196–198], although some studies have reported discrepancies with conventional laboratory determinations [199]. The clinical utility of the parameters measured by portable coagulometers may therefore be limited compared to more comprehensive viscoelastic assays.
Viscoelastic methods offer a rapid and dynamic assessment of coagulation to support timely clinical decision-making, leading to increasing confidence and wider adoption of these techniques [200, 201]. Case series utilizing viscoelastic testing in trauma patients have been published, including studies employing rotational thromboelastography in trauma patients [179]. Johansson et al. [180] implemented a hemostatic resuscitation protocol incorporating early platelet and fresh frozen plasma (FFP) transfusion guided by thrombelastography in a before-and-after study, demonstrating improved patient outcomes. In a retrospective study of cardiovascular surgery patients, the combined use of thromboelastometry and portable coagulometry resulted in a reduction in blood product transfusion and thromboembolic events, although mortality was not significantly affected [202]. Rapid thrombelastography, a newer variant of viscoelastic testing initiated by kaolin and tissue factor, appears to offer reduced measurement time compared to conventional thrombelastography [203].
Despite the increasing adoption of viscoelastic methods, their clinical utility remains debated. A recent Cochrane systematic review by Hunt et al. [204] found limited evidence supporting the accuracy of thrombelastography and thromboelastometry in diagnosing trauma-induced coagulopathy, and therefore could not provide definitive recommendations regarding their use [204]. Another systematic review by Da Luz et al. [205] concluded that while limited evidence from observational studies supports the use of viscoelastic tests for diagnosing early traumatic coagulopathy and predicting blood product transfusion, their impact on mortality and other patient-important outcomes remains unclear [205]. Several limitations of viscoelastic methods have been described. Larsen et al. [206] found that thrombelastography struggled to differentiate coagulopathies caused by dilution from those caused by thrombocytopenia, while thromboelastometry demonstrated better discrimination and guided appropriate treatment strategies [206]. Thrombelastography may potentially lead to unnecessary platelet transfusions, whereas thromboelastometry may facilitate goal-directed fibrinogen substitution. While viscoelastic method utilization is rapidly expanding, controversy persists regarding their overall utility in detecting and managing post-traumatic coagulopathy.
The level of agreement between viscoelastic methods and standard coagulation tests also remains a subject of debate. While some studies report acceptable agreement [207–209], others have found significant discrepancies [25, 199, 210, 211], even between different viscoelastic methods (thrombelastography and thromboelastometry). Hagemo et al. [212] found that the correlation varied substantially at different stages of clot formation and between different centers, highlighting the need for standardization and clarification of these techniques. A limitation of viscoelastic tests is their reduced sensitivity in detecting and monitoring platelet dysfunction caused by antiplatelet drugs. In cases where platelet dysfunction is suspected, point-of-care platelet function tests, such as whole blood impedance aggregometry, should be used in conjunction with viscoelastic tests [213, 214]. Further research is needed to refine the optimal integration of viscoelastic and conventional coagulation monitoring in trauma care, and clinicians should utilize their clinical judgment in developing local policies while awaiting further evidence.
While theoretically plausible that changes in coagulation measures like D-dimers might help identify patients with ongoing bleeding, evidence for their clinical utility in this context is limited. A single publication reported a low positive predictive value of D-dimers (1.8%) in postoperative and post-traumatic settings [215]. Therefore, traditional methods for detecting ongoing bleeding, such as serial clinical evaluations and radiology (ultrasound, CT, or angiography), remain the cornerstone of clinical assessment.
Tissue Oxygenation, Type of Fluid, and Temperature Management
Tissue Oxygenation
Recommendation 13
We recommend a target systolic blood pressure of 80–90 mmHg until major bleeding has been stopped in the initial phase following trauma without brain injury. (Grade 1C)
In patients with severe TBI (GCS ≤8), we recommend that a mean arterial pressure ≥80 mmHg be maintained. (Grade 1C)
Restricted Volume Replacement
Recommendation 14
We recommend use of a restricted volume replacement strategy to achieve target blood pressure until bleeding can be controlled. (Grade 1B)
Vasopressors and Inotropic Agents
Recommendation 15
In the presence of life-threatening hypotension, we recommend administration of vasopressors in addition to fluids to maintain target arterial pressure. (Grade 1C)
We recommend infusion of an inotropic agent in the presence of myocardial dysfunction. (Grade 1C)
Rationale
Traditional trauma resuscitation strategies have emphasized early and aggressive fluid administration to restore blood volume and maintain tissue oxygenation. However, this approach can paradoxically exacerbate bleeding by increasing hydrostatic pressure at the injury site, disrupting clot formation, diluting coagulation factors, and inducing hypothermia. The concept of “damage control resuscitation” aims to achieve a lower-than-normal blood pressure, termed “permissive hypotension,” to mitigate these adverse effects of aggressive fluid resuscitation while acknowledging the potential for transient tissue hypoperfusion [216]. The overall effectiveness of permissive hypotension in improving outcomes remains under investigation in ongoing randomized clinical trials.
However, two studies published in the 1990s suggested increased survival with low and delayed fluid volume resuscitation in penetrating trauma [217] and combined penetrating and blunt trauma [218]. In contrast, two subsequent trials in patients with penetrating and blunt trauma [219] or blunt trauma alone [220] failed to demonstrate significant survival differences with restrictive fluid strategies.
Several retrospective analyses in recent years have indicated that aggressive resuscitation techniques, often initiated pre-hospitally, may be detrimental to trauma patients [9, 28, 221, 222]. One study linked aggressive fluid resuscitation to an increased risk of secondary abdominal compartment syndrome (ACS) in patients with severe extremity injuries [221], identifying early large-volume crystalloid administration as the strongest predictor of secondary ACS. Another retrospective analysis using the German Trauma Registry, encompassing over 17,000 multiply injured patients, demonstrated that the incidence of coagulopathy increased with increasing volumes of pre-clinical intravenous fluid administration [9]. Coagulopathy rates exceeded 40% in patients receiving >2000 ml, 50% in those receiving >3000 ml, and 70% in those receiving >4000 ml. Using the same registry, a matched-pairs analysis (n=1896) showed that multiply injured trauma patients with an Injury Severity Score (ISS) ≥16 and systolic blood pressure ≥60 mmHg at the scene who received low-volume resuscitation (0–1500 ml) had higher survival rates compared to those receiving high-volume resuscitation (≥1501 ml) [28]. These findings are supported by a retrospective analysis of the US National Trauma Data Bank [222], which analyzed over 776,000 patients and found that pre-hospital intravenous fluid administration was associated with higher mortality, particularly in patients with penetrating injuries, hypotension, severe head injury, and those requiring immediate surgery. The authors concluded that routine pre-hospital intravenous fluid administration for all trauma patients should be discouraged, although this conclusion has been debated [223].
Initial use of a restrictive volume replacement strategy is supported by a prospective randomized trial evaluating hypotensive resuscitation in trauma patients with hemorrhagic shock [224]. In this study, primarily involving patients with penetrating trauma, those with at least one documented in-hospital systolic blood pressure ≤90 mmHg were randomized to target minimum mean arterial pressures of 50 mmHg or 65 mmHg. While no statistically significant difference in actual mean arterial pressure was observed between the groups, patients in the lower target MAP group experienced reduced 24-hour postoperative death and coagulopathy, along with lower overall fluid and blood product transfusion requirements. Another study by Brown et al. [225] in over 1200 trauma patients with ISS >15, approximately half of whom were hypotensive, showed that pre-hospital crystalloid administration >500 ml was associated with worse outcomes in patients without pre-hospital hypotension, but not in hypotensive patients, suggesting that pre-hospital volume resuscitation should be goal-directed, guided by the presence or absence of hypotension. Recently, Schreiber et al. [226] assessed the feasibility and safety of controlled resuscitation (n=97) versus standard resuscitation (n=95) in hypotensive trauma patients, randomizing patients in the pre-hospital setting. Eligible patients had pre-hospital systolic blood pressure ≤90 mmHg. Controlled resuscitation patients received 250 ml fluid if no radial pulse or SAP <70 mmHg, while standard resuscitation patients received 1-2 liters of crystalloid. The controlled resuscitation group received significantly less pre-hospital fluid and demonstrated a trend towards improved survival, although the study was not powered to detect a statistically significant survival difference.
A meta-analysis by Kwan et al. [227] analyzing randomized trials investigating timing and volume of intravenous fluid administration in bleeding trauma patients identified three trials addressing timing (total of 1957 patients) and three trials investigating volume load (only 171 patients). This meta-analysis failed to demonstrate an advantage associated with delayed versus early fluid administration or smaller versus larger fluid volumes in this limited set of prospective studies. Another meta-analysis assessing seven retrospective observational studies (13,687 patients) and three prospective studies (798 patients) suggested a small benefit favoring restricted volume replacement [228], but cautioned about the high risk of selection bias and clinical heterogeneity in the available studies.
It is crucial to note that damage control resuscitation strategies using restrictive volume replacement are contraindicated in patients with TBI and spinal cord injuries. Adequate cerebral perfusion pressure is paramount to ensure oxygen delivery to the injured central nervous system in these patients [229]. Rapid bleeding control is particularly critical in TBI and spinal cord injury patients. Furthermore, permissive hypotension should be carefully considered and may be contraindicated in elderly patients, especially those with pre-existing chronic arterial hypertension [230].
In conclusion, while supported by emerging literature, particularly retrospective analyses and smaller prospective trials, the evidence base for a damage control resuscitation strategy targeting a lower-than-normal systolic blood pressure of 80–90 mmHg with restricted fluid replacement in patients without TBI or spinal injury remains limited, with a lack of robust evidence from large-scale RCTs.
Vasopressors may be necessary as a temporizing measure to sustain life and maintain tissue perfusion in life-threatening hypotension, even during ongoing fluid resuscitation and before hypovolemia is fully corrected. Norepinephrine (NE) is frequently used to restore arterial pressure in septic and hemorrhagic shock and is now recommended as the vasopressor of choice in septic shock [231]. While NE has some β-adrenergic effects, its predominant action is vasoconstriction mediated by α-adrenergic stimulation. Arterial α-adrenergic stimulation increases arterial resistance and cardiac afterload, while NE exerts both arterial and venous α-adrenergic stimulation [232]. In addition to its arterial vasoconstrictor effects, NE induces venoconstriction, particularly in the splanchnic circulation, increasing pressure in capacitance vessels and actively shifting splanchnic blood volume into the systemic circulation [233]. This venous adrenergic stimulation can mobilize some blood from the unstressed venous volume, the volume filling blood vessels without generating intravascular pressure. Furthermore, β2-adrenergic receptor stimulation can decrease venous resistance and increase venous return [233].
Animal studies investigating uncontrolled hemorrhage have suggested that NE infusion reduces fluid resuscitation requirements to achieve a given arterial pressure target, is associated with lower blood loss, and significantly improves survival [234, 235]. However, the effects of NE in human hemorrhagic shock have not been rigorously investigated in prospective clinical trials. An interim analysis from an ongoing multicenter prospective cohort study suggested that early vasopressor use for hemodynamic support after hemorrhagic shock may be deleterious compared to aggressive volume resuscitation and should be approached cautiously [236]. However, this study had limitations, including being a secondary analysis not designed to specifically address the hypothesis tested and a higher rate of thoracotomy in the vasopressor group. Prospective studies are needed to definitively define the role of vasopressors in human hemorrhagic shock.
A double-blind randomized trial assessing the safety and efficacy of adding vasopressin to resuscitative fluids has been conducted [237]. Patients received either fluid alone or fluid plus vasopressin (bolus 4 IU followed by i.v. infusion of 2.4 IU/h for 5 hours). The fluid plus vasopressin group required significantly less total resuscitation fluid volume over 5 days compared to the control group (P = 0.04). Adverse event rates, organ dysfunction, and 30-day mortality were similar between groups.
Vasopressors, such as norepinephrine, may be valuable as transient life-sustaining measures to maintain arterial pressure and tissue perfusion in the face of life-threatening hypotension. If used, it is essential to adhere to recommended systolic blood pressure targets (80–90 mmHg) in patients without TBI.
Because vasopressors can increase cardiac afterload, particularly with excessive infusion rates or pre-existing impaired left ventricular function, assessing cardiac function during initial ultrasound examination is crucial. Cardiac dysfunction may be present in trauma patients due to cardiac contusion, pericardial effusion, or secondary to brain injury with intracranial hypertension. Myocardial dysfunction necessitates treatment with an inotropic agent, such as dobutamine or epinephrine. In the absence of cardiac function assessment or cardiac output monitoring, which is often the case in the early phases of hemorrhagic shock management, cardiac dysfunction should be suspected in patients exhibiting poor response to both fluid expansion and norepinephrine administration.
Type of Fluid
Recommendation 16
We recommend that fluid therapy using isotonic crystalloid solutions be initiated in the hypotensive bleeding trauma patient. (Grade 1A)
We suggest that excessive use of 0.9 % NaCl solution be avoided. (Grade 2C)
We recommend that hypotonic solutions such as Ringer’s lactate be avoided in patients with severe head trauma. (Grade 1C)
We suggest that the use of colloids be restricted due to the adverse effects on hemostasis. (Grade 2C)
Rationale
While fluid resuscitation represents the initial and essential step in restoring tissue perfusion in severe hemorrhagic shock, the optimal type of fluid—crystalloids or colloids, and the specific crystalloid or colloid solution—for initial resuscitation of bleeding trauma patients remains unclear.
In most trauma studies, 0.9% sodium chloride (normal saline) has been the predominant crystalloid solution used. However, recent evidence suggests that normal saline may contribute to increased acidosis and a higher incidence of kidney injury in both healthy volunteers and critically ill adults [238, 239]. In contrast, balanced electrolyte solutions are formulated to contain electrolyte concentrations that are physiologically similar or near-physiological to human plasma. A small prospective randomized trial in 46 trauma patients demonstrated that a balanced electrolyte solution improved acid-base status and resulted in less hyperchloremia at 24 hours post-injury compared to 0.9% sodium chloride [240]. A secondary analysis of this study further suggested that the use of a balanced electrolyte solution offered a net cost benefit compared to 0.9% saline [241]. Therefore, if 0.9% sodium chloride is utilized for resuscitation, its use should be limited to a maximum volume of 1–1.5 liters.
When crystalloids are employed, hypotonic solutions such as Ringer’s lactate should be avoided, particularly in patients with TBI, to minimize potential fluid shifts into damaged cerebral tissue and exacerbate cerebral edema. Furthermore, solutions with the potential to restore physiological pH may be advantageous. A recent study demonstrated that Ringer’s acetate solution more rapidly improved splanchnic dysoxia, as measured by gastric tonometry, compared to Ringer’s lactate [242]. Whether certain isotonic balanced crystalloids offer a definitive advantage in terms of reduced morbidity or mortality remains uncertain and requires further evaluation [241, 243].
The most recent Cochrane meta-analysis examining the comparative effectiveness of colloids versus crystalloids for fluid resuscitation in critically ill patients failed to demonstrate that colloids reduce the risk of death compared to crystalloid resuscitation in ICU settings [244]. This analysis, comparing albumin or plasma protein fraction with crystalloids across 24 trials involving nearly 10,000 patients, found a pooled risk ratio (RR) of 1.01 (95% CI 0.93 to 1.10). Twenty-five trials comparing hydroxyethyl starch (HES) to crystalloids (totaling over 9,000 patients) suggested a potentially harmful effect of HES [RR 1.10 (1.02 to 1.19)]. Modified gelatin, assessed in 11 trials (506 patients), showed neither benefit nor harm [RR 0.91 (0.49 to 1.72)]. The authors concluded that current evidence does not support a survival benefit with colloid resuscitation, and HES may even be detrimental. However, this meta-analysis did not specifically analyze or discuss the timing, duration, or dosages of fluid resuscitation in the context of trauma. Nevertheless, robust evidence supporting the superiority of colloids over crystalloids in trauma resuscitation remains lacking.
Given that colloids are also generally more expensive than crystalloids, and considering the restricted volume replacement strategy recommended during initial trauma management, crystalloids appear to be a justified initial fluid choice for hypotensive bleeding trauma patients. Even in later resuscitation stages, large-volume crystalloid administration is not independently associated with multiple organ failure [245]. Additionally, in scenarios where high ratios of FFP:RBC transfusion are not feasible, retrospective data suggest that resuscitation with at least 1 liter of crystalloid per unit of RBC may be associated with reduced overall mortality [246].
Currently, the optimal colloid solution, if any, to use when crystalloids fail to restore target blood pressure remains unclear. A Cochrane meta-analysis comparing different colloid solutions in over 5,000 patients requiring volume replacement [247] found no evidence indicating that any specific colloid solution is demonstrably more effective or safer than others, although wide confidence intervals preclude ruling out clinically significant differences between colloids. However, meta-analyses have presented conflicting data, with some suggesting increased kidney injury and mortality with HES [248, 249] and others finding no differences in death or acute kidney injury in surgical patients receiving 6% HES [250]. The generalizability of these studies, often conducted in clinical contexts distinct from acute hypovolemic trauma, remains questionable. In vitro studies have also suggested that all HES and gelatin solutions can impair coagulation and platelet function [251], with gelatin-induced coagulopathy potentially more readily reversible with fibrinogen administration than HES-induced coagulopathy. Only one small RCT has suggested a benefit for HES solutions in trauma patients, with HES (130/0.4) demonstrating improved lactate clearance and reduced renal injury compared to saline in 67 penetrating trauma patients [252]. Blunt trauma patients in this study showed no significant differences between solutions. If colloids are considered for patients in whom crystalloids fail to achieve target blood pressure, dosing should adhere to recommended limits, and if HES is employed, modern HES solutions should be preferred.
Hypertonic solutions have also been investigated in trauma resuscitation. A double-blind RCT in 209 blunt trauma patients comparing 7.5% hypertonic saline with 6% dextran 70 to lactated Ringer’s solution found no significant difference in organ failure or ARDS-free survival in the intent-to-treat analysis [253]. However, a subgroup analysis suggested improved ARDS-free survival in patients requiring ≥10 units of packed RBCs [253]. Clinical trials in brain injury patients have shown that hypertonic saline more effectively reduces intracranial pressure compared to dextran solutions or 20% mannitol [254]. However, Cooper et al. [255] found minimal difference in neurological function at 6 months after TBI in patients receiving pre-hospital hypertonic saline versus conventional fluid resuscitation. Two large prospective randomized multicenter studies by Bulger and colleagues [256, 257], involving over 2,100 patients, failed to demonstrate any advantage of out-of-hospital hypertonic fluids compared to 0.9% saline in neurological outcomes after severe TBI or survival after traumatic hypovolemic shock. Conversely, a recent study suggested that hypertonic solutions may interfere with coagulation in trauma patients [258].
In conclusion, current evidence suggests that hypertonic saline solutions are generally safe but do not consistently improve survival or neurological outcomes after TBI. Only one study has reported improved survival with hypertonic saline dextran compared to normal saline [259].
Erythrocytes
Recommendation 17
We recommend a target Hb of 7 to 9 g/dl. (Grade 1C)
Rationale
Oxygen delivery to tissues, a critical determinant of cellular function and survival, is a product of blood flow and arterial oxygen content, directly related to hemoglobin (Hb) concentration. While intuitively, decreasing Hb concentration might be expected to induce tissue hypoxia, the physiological response to acute normovolemic anemia is complex, involving compensatory mechanisms at both macro- and microcirculatory levels to maintain oxygen delivery.
Randomized controlled trials (RCTs) evaluating Hb thresholds for red blood cell (RBC) transfusion in critically ill patients have consistently demonstrated that restrictive transfusion strategies (Hb thresholds between 7 and 9 g/dL) are as safe, or potentially safer, than liberal transfusion strategies (thresholds ≥9 g/dL) [260–263], with the possible exceptions of patients undergoing cardiac surgery [264] or experiencing acute coronary syndromes. Critically, these studies have generally excluded patients with active massive bleeding. No prospective RCTs have directly compared restrictive and liberal transfusion regimens specifically in trauma patients. However, a subset analysis of 203 trauma patients from the Transfusion Requirements in Critical Care (TRICC) trial [260] suggested that a restrictive transfusion regimen (Hb transfusion trigger <7 g/dL) was safe for resuscitated and critically ill trauma patients [265].
Observational studies in trauma patients have consistently shown that liberal RBC transfusion strategies are associated with increased morbidity and mortality [266–268], infectious complications [269, 273, 274], lung injury [270–272], and renal failure [269].
Given the potential for anemia to contribute to secondary ischemic damage, concerns have been raised regarding the safety of restrictive transfusion strategies specifically in patients with TBI. Early clinical data were primarily derived from retrospective observational studies with methodological limitations, yielding inconsistent results regarding the effects of RBC transfusion on cerebral perfusion and metabolism markers in isolated severe TBI. Two systematic reviews published in 2012 emphasized the lack of high-level evidence supporting specific Hb transfusion triggers in this population [275, 276].
More recently, two studies have focused on the impact of anemia and RBC transfusion on neurological outcomes after TBI [277, 278]. A retrospective review of prospectively collected data from over 1,100 patients with GCS ≤8 without hemorrhagic shock found that RBC transfusion was associated with worse outcomes (28-day survival, ARDS-free survival, 6-month neurological outcome) when initial Hb was >10 g/dl [277]. No relationship between RBC transfusion and outcomes was observed in patients with initial Hb ≤10 g/dl [277]. In a 2×2 factorial design RCT of 200 TBI patients, Robertson et al. compared two Hb transfusion thresholds (7 or 10 g/dl) and erythropoietin (EPO) administration versus placebo [278]. No advantage was found for the 10 g/dl Hb threshold. In the 7 g/dl threshold group, 42.5% of patients had favorable neurological outcomes at 6 months, compared to 33.0% in the 10 g/dl threshold group (95% CI for difference −0.06 to 0.25). No difference in mortality was observed, but more thromboembolic events occurred in the 10 g/dl threshold group [278]. Taken together, current evidence suggests that patients with severe TBI should not be managed with a Hb transfusion threshold different from other critically ill patients. A restrictive transfusion strategy targeting a Hb of 7-9 g/dL is generally considered safe and appropriate.
Erythrocytes contribute to hemostasis through multiple mechanisms, including influencing platelet biochemical and functional responsiveness, promoting platelet margination through rheological effects, and supporting thrombin generation [279]. The precise effects of Hct on blood coagulation are complex and not fully elucidated [280]. Acute Hct reduction can increase bleeding time [281, 282], reversible with re-transfusion [281]. This may relate to elastase on RBC membranes, which can activate coagulation factor IX [283, 284]. However, animal models have shown that moderate Hct reduction does not increase blood loss from spleen injury [282], and isolated in vitro Hct reduction did not compromise blood coagulation as assessed by thromboelastometry [285].
Alternative strategies to raise Hb levels beyond RBC transfusion have been explored. The erythropoietic response is often blunted in trauma patients [286], making erythropoietin (EPO) administration a potentially attractive option. An early prospective randomized trial in ICU patients (48% trauma patients) showed a significant reduction in RBC transfusion rates with EPO [287], and a subgroup analysis suggested reduced 28-day mortality in trauma patients [287]. However, a subsequent larger prospective randomized trial in ICU patients (54% trauma patients) did not replicate the reduction in RBC transfusions [288]. Thrombotic complications were higher in EPO-treated patients, particularly in those without heparin prophylaxis [288]. Nevertheless, a trend towards reduced mortality was observed in the overall ICU population, and trauma patients showed lower 29-day and 140-day mortality with EPO [288]. A third prospective randomized trial in patients with major blunt orthopedic trauma [289] found no significant effect of EPO, although a high drop-out rate in the study may have contributed to this non-significant result.
The limited effect of EPO on transfusion needs may be attributed to altered iron metabolism after trauma, rendering iron less available for erythropoiesis [286]. Intravenous iron administration, with [290, 291] or without [292] concomitant EPO, has shown promise in reducing RBC transfusions [290–292], postoperative infections [290–292], length of hospital stay [291], and mortality in hip fracture patients [291]. While intravenous iron appears promising, oral iron is largely ineffective in this setting [293]. Ongoing research, such as the PAHFRAC-01 project [294], aims to further clarify the benefits of intravenous iron and EPO in hip fracture patients.
Meta-analyses in non-trauma surgical patients suggest that preoperative intravenous iron administration can correct preoperative anemia and reduce RBC transfusion rates in elective surgery, but may also increase infection risk [295]. This potential risk remains to be fully evaluated for postoperative intravenous iron administration and in trauma patients specifically. Interestingly, intravenous iron treatment in hemodialysis patients was associated with a trend towards lower infection rates, mortality, and shorter hospital stays [296]. Similarly, intravenous iron treatment in septic anemic mice did not increase mortality and effectively corrected anemia [297]. Short-term preoperative treatment with iron carboxymaltose and EPO also significantly reduced postoperative infectious complications and shortened hospital stays in anemic orthopedic surgery patients [291], with reduced 30-day mortality observed in hip fracture patients [291]. The potential adverse effects of intravenous iron administration in trauma patients may be overestimated and warrant further investigation.
Temperature Management
Recommendation 18
We recommend early application of measures to reduce heat loss and warm the hypothermic patient in order to achieve and maintain normothermia. (Grade 1C)
Rationale
Hypothermia, defined as a core body temperature below 35 °C, is a common and serious complication in trauma patients, particularly those with severe injuries and shock. Hypothermia is a well-recognized component of the “lethal triad” of trauma, along with coagulopathy and acidosis [298–300]. Body temperatures below 34°C significantly compromise blood coagulation. While this coagulopathic effect is observed when coagulation tests (PT and APTT) are performed at low temperatures, it is not consistently reflected when tests are conducted at the standard 37°C laboratory temperature, potentially masking the true extent of hypothermia-induced coagulopathy in routine clinical assessments.
The profound clinical effects of hypothermia contribute to increased morbidity and mortality [301]. Hypothermic patients often require larger volumes of blood products [302]. A retrospective study of over 600 trauma patients requiring massive transfusion demonstrated that a temperature below 34°C was associated with a greater than 80% increased independent risk of mortality, even after adjusting for differences in shock severity, coagulopathy, injury severity, and transfusion requirements [303]. A recent secondary data analysis of over 11,000 patients with severe TBI from the Pennsylvania Trauma Outcome Study (PTOS) revealed that spontaneous hypothermia at hospital admission was associated with a significantly increased risk of mortality in this patient population [304].
Strategies to prevent hypothermia and mitigate hypothermia-induced coagulopathy include removing wet clothing, covering the patient to minimize further heat loss, increasing ambient temperature, utilizing forced air warming devices, administering warmed intravenous fluids, and, in extreme cases, employing extracorporeal re-warming devices [305–307].
While accidental or induced hypothermia should be avoided in trauma patients without TBI, the role of therapeutic hypothermia in patients with TBI remains a subject of ongoing investigation and debate. Several large multicenter clinical trials have failed to demonstrate a clear benefit of therapeutic hypothermia in TBI [308–310]. However, a recent meta-analysis by Crossley et al., incorporating single-center studies, suggested an overall reduction in mortality and poor outcomes [311]. Earlier meta-analyses examining mortality and neurological outcomes associated with mild hypothermia in TBI did not consistently show such benefits, potentially due to variations in inclusion/exclusion criteria and study methodologies [312, 313]. Differences in the speed of hypothermia induction and duration of cooling may also contribute to the conflicting results. For example, prolonged mild therapeutic hypothermia (5 days) may be more effective than short-term cooling (2 days) in controlling refractory intracranial hypertension in adults with severe TBI [314, 315]. Furthermore, the comparative effectiveness of hypothermia may depend on whether it is compared to standard treatment allowing fever episodes or to strict temperature control within a narrow normothermic range (35.5–37°C) [310]. Currently, no definitive recommendation can be made in favor of therapeutic whole-body hypothermia in TBI patients.
Rapid Control of Bleeding
Damage Control Surgery
Recommendation 19
We recommend that damage control surgery be employed in the severely injured patient presenting with deep hemorrhagic shock, signs of ongoing bleeding and coagulopathy. (Grade 1B)
Other factors that should trigger a damage control approach are severe coagulopathy, hypothermia, acidosis, inaccessible major anatomic injury, a need for time-consuming procedures or concomitant major injury outside the abdomen. (Grade 1C)
We recommend primary definitive surgical management in the hemodynamically stable patient and in the absence of any of the factors above. (Grade 1C)
Rationale
Severely injured patients arriving at the hospital with ongoing bleeding or profound hemorrhagic shock face a grim prognosis without rapid bleeding control, effective resuscitation, and blood product transfusion. This is especially true for patients with uncontrolled hemorrhage from multiple penetrating injuries or major abdominal injuries combined with unstable pelvic fractures, leading to bleeding from fracture sites and retroperitoneal vessels. The ultimate consequence in these scenarios is the exhaustion of physiological reserves, culminating in profound acidosis, hypothermia, and coagulopathy—the “bloody vicious cycle” or “lethal triad.”
In 1983, Stone et al. described the technique of abbreviated laparotomy, incorporating packing to control hemorrhage and deferring definitive surgical repair until coagulation is restored [316]. Since then, numerous publications have documented the benefits of this staged approach, now widely known as “damage control surgery” [317–320]. Damage control surgery should be considered in patients with major abdominal injuries requiring adjunctive angioembolization, major abdominal injuries requiring early evaluation of other injuries, or major abdominal injuries accompanied by traumatic limb amputation. Intraoperatively, factors that should trigger a damage control approach include core temperature ≤34°C, pH ≤7.2, inaccessible major venous injury, need for time-consuming procedures in a patient with suboptimal response to resuscitation, or inability to achieve hemostasis due to refractory coagulopathy [321, 322].
Damage control surgery for abdominal trauma comprises three key phases. The first phase involves an abbreviated resuscitative laparotomy focused on rapid hemorrhage control, restoration of blood flow as needed, and contamination control. This phase prioritizes speed and avoids time-consuming definitive organ repairs, which are deferred to a later stage. The abdomen is packed with surgical sponges to tamponade bleeding, and temporary abdominal closure is performed. Abdominal packing aims to compress liver ruptures, exert direct pressure on bleeding sources, and facilitate subsequent hemostasis through angiography and/or correction of the lethal triad. Pack removal is ideally delayed for at least 48 hours to minimize the risk of re-bleeding.
The second phase of damage control surgery is intensive care unit (ICU) management, focusing on core re-warming, correction of acid-base imbalances and coagulopathy, and optimizing ventilation and hemodynamic status. If angiography or further injury investigation is needed, it is performed during this phase.
The third phase is definitive surgical repair, undertaken only when target physiological parameters have been achieved [95, 317–320, 323, 324]. While the conceptual basis of “damage control” is intuitively sound, definitive evidence from RCTs is lacking. Retrospective studies, however, support the concept, demonstrating reduced morbidity and mortality rates in carefully selected patient populations [320].
The principles of damage control have been extended to orthopedic injuries in severely injured patients, termed “damage control orthopedics” by Scalea et al. [325]. Relevant fractures are initially stabilized with external fixators rather than pursuing primary definitive osteosynthesis [325–327]. This less traumatic and shorter surgical procedure aims to minimize secondary procedure-related trauma. Definitive osteosynthesis surgery can be performed 4–14 days later, when the patient’s physiological status has improved sufficiently. Retrospective clinical studies and prospective cohort studies generally support the damage control orthopedics approach. A single available randomized study suggests a benefit for this strategy in “borderline” patients [327]. The damage control concept has also been adapted for thoracic and neurosurgical injuries [328, 329]. Beyond surgical techniques, damage control anesthesia and resuscitation encompass a range of critical measures outlined in other recommendations within these guidelines.
Pelvic Ring Closure and Stabilization
Recommendation 20
We recommend that patients with pelvic ring disruption in hemorrhagic shock undergo immediate pelvic ring closure and stabilization. (Grade 1B)
Packing, Embolization, and Surgery
Recommendation 21
We recommend that patients with ongoing hemodynamic instability despite adequate pelvic ring stabilization receive early pre-peritoneal packing, angiographic embolization and/or surgical bleeding control. (Grade 1B)
Rationale
Mortality rates for patients with severe pelvic ring disruptions and hemodynamic instability remain unacceptably high [330, 331]. Early recognition of these injuries and immediate interventions to reduce pelvic ring disruption, stabilize the pelvis, and control bleeding are therefore critical. Markers indicative of pelvic hemorrhage include anterior-posterior and vertical shear deformities on standard radiographs, CT “blush” (active arterial extravasation), elevated bladder compression pressure, pelvic hematoma visualized on CT, and persistent hemodynamic instability despite adequate fracture stabilization [332–334].
Initial management of pelvic fractures focuses on controlling venous and cancellous bone bleeding through pelvic ring closure as the primary step [335]. Some institutions primarily utilize external fixators for hemorrhage control from pelvic fractures [332], but pelvic closure can also be achieved using pelvic binders, pelvic C-clamps, or improvised methods like bed sheets [335, 336]. In addition to pelvic closure, fracture stabilization, and the tamponade effect of hematoma, pre-peritoneal, extraperitoneal, or retroperitoneal packing can reduce or arrest venous bleeding [337–339]. Pre-peritoneal packing is used to decrease the need for pelvic embolization and can be performed concurrently with, or shortly after, initial pelvic fracture stabilization. Pelvic embolization targets arterial bleeding and is the most frequently embolized vascular bed in trauma, making it the most extensively studied [340]. Pelvic packing can facilitate early intrapelvic bleeding control, providing crucial time for more selective hemorrhage management strategies [337, 339].
Resuscitative endovascular balloon occlusion of the aorta (REBOA) has been utilized in patients with end-stage shock following blunt and penetrating trauma, often in conjunction with embolization of pelvic vasculature. Reports on REBOA in trauma are limited, with a paucity of published clinical trials. Combined approaches incorporating REBOA are evolving [331]. These techniques can be integrated with subsequent laparotomy if deemed necessary [337]. This combined approach may potentially improve the historically high mortality rates observed in patients with major pelvic injuries undergoing laparotomy as the primary intervention. However, non-therapeutic laparotomy should be avoided [341]. Time to pelvic embolization for hemodynamically unstable pelvic fractures may directly impact survival [331, 342].
Angiography and embolization are now widely accepted as highly effective methods for controlling arterial bleeding that is refractory to fracture stabilization [146, 332, 336, 339, 341, 343, 344]. Radiological management can also be effectively applied to abdominal and thoracic bleeding [345–349]. Martinelli et al. [350] reported the use of intra-aortic balloon occlusion to reduce bleeding and facilitate transport to the angiography suite. Conversely, Morozumi et al. [351] advocate for the use of mobile digital subtraction angiography within the emergency department, enabling arterial embolization to be performed directly by trauma surgeons. Several authors suggest that permissive hypotension, coupled with pelvic stabilization and/or angiography (damage control resuscitation, hypertonic solutions, controlled hypothermia), can improve survival outcomes. Institutional variations in the capacity for timely angiography and embolization may contribute to the diverse treatment algorithms proposed in the literature. Reports of 100% higher mortality associated with transcatheter angiographic embolization performed during off-hours, attributed to limited radiological service availability [352], underscore the necessity of a multidisciplinary approach to managing these severe injuries.
Local Hemostatic Measures
Recommendation 22
We recommend the use of topical hemostatic agents in combination with other surgical measures or with packing for venous or moderate arterial bleeding associated with parenchymal injuries. (Grade 1B)
Rationale
A broad spectrum of topical hemostatic agents is available as adjuncts to conventional surgical techniques for achieving hemorrhage control. These agents are particularly useful when surgical access to the bleeding site is challenging. Topical hemostatic agents encompass collagen, gelatin, or cellulose-based products, fibrin sealants, and synthetic glues or adhesives, suitable for both external and internal bleeding. Polysaccharide-based and inorganic hemostatics are predominantly used and approved for external bleeding control.
The selection of topical hemostatic agents should consider several factors, including the specific surgical procedure, cost, severity of bleeding, patient coagulation status, and the unique characteristics of each agent. Certain agents should be avoided when autotransfusion is planned, and other specific contraindications should be carefully considered [353, 354]. The hemostatic efficacy of these agents was initially assessed in animal models, but growing clinical experience in human applications is now available [353–369].
Different classes of local hemostatic agents can be broadly categorized based on their composition and hemostatic mechanisms:
- Collagen-based hemostats: These agents, typically bovine-derived, promote platelet aggregation and activation upon contact with blood. Examples include collagen sponges and fleeces.
- Gelatin-based hemostats: Derived from porcine or bovine gelatin, these agents provide a matrix for clot formation. Examples include gelatin sponges and flowable gelatin.
- Cellulose-based hemostats: Oxidized regenerated cellulose (ORC) products, such as Surgicel, promote clot formation by providing a scaffold and activating the intrinsic coagulation pathway.
- Fibrin sealants: These agents mimic the final stages of the coagulation cascade, delivering fibrinogen and thrombin to the bleeding site to form a fibrin clot. Examples include Evicel and Tisseel.
- Thrombin-based hemostats: These agents deliver topical thrombin, accelerating the conversion of fibrinogen to fibrin and promoting rapid clot formation.
- Cyanoacrylate adhesives: Synthetic adhesives like Dermabond create a strong, waterproof seal for skin closure and can be used for superficial bleeding.
- Polysaccharide-based hemostats: These agents, such as HemCon bandage (chitosan-based), promote hemostasis through mucoadhesion and interaction with red blood cells. Primarily used for external bleeding.
- Inorganic hemostats: Kaolin-based hemostats (e.g., QuikClot) and zeolite-based hemostats promote hemostasis by concentrating coagulation factors and platelets at the bleeding site. Primarily used for external bleeding.
While evidence supporting the use of topical hemostatic agents is largely observational, these agents have become widely adopted as valuable adjuncts in surgical hemostasis.
Initial Management of Bleeding and Coagulopathy
Coagulation Support
Recommendation 23
We recommend that monitoring and measures to support coagulation be initiated immediately upon hospital admission. (Grade 1B)
Rationale
Several tools have been developed to evaluate trauma-related coagulopathy [370], largely confirming the underlying pathophysiological mechanisms described previously [371, 372]. While general pathophysiological mechanisms contributing to trauma-induced coagulopathy are well-recognized, it is crucial to rapidly determine the specific type and severity of coagulopathy in each individual patient to guide targeted therapy [373].
Early monitoring of coagulation is essential to detect trauma-induced coagulopathy and identify its primary drivers, including hyperfibrinolysis [13, 25, 179, 183, 374]. Early therapeutic intervention can improve coagulation test results [375], reduce the need for RBC, FFP, and platelet transfusions [12, 376], decrease post-traumatic multi-organ failure incidence, shorten hospital length of stay [12], and potentially improve survival [377, 378]. Interestingly, the success of early algorithm-based and goal-directed coagulation management in reducing transfusions and improving outcomes, including mortality, has also been demonstrated in cardiac surgery [202, 379–381]. Early algorithm-based and goal-directed coagulation management is therefore likely to improve outcomes for severely injured patients [382, 383]. This has been supported by prospective randomized trials [384] and large-scale implementation studies [385]. However, some studies have not shown a survival benefit [375, 386, 387]; variations in results may be attributed to the choice of coagulation monitoring tests (negative trials often used conventional laboratory values like PT, APTT, and platelet count) and the type of therapy used (negative trials frequently relied solely on FFP and platelets [379–381, 384]).
Initial Coagulation Resuscitation
Recommendation 24
In the initial management of patients with expected massive hemorrhage, we recommend one of the two following strategies:
- Plasma (FFP or pathogen-inactivated plasma) in a plasma–RBC ratio of at least 1:2 as needed. (Grade 1B)
- Fibrinogen concentrate and RBC according to Hb level. (Grade 1C)
Rationale
“Initial resuscitation” is defined here as the period from emergency department arrival until coagulation monitoring results (coagulation screen, fibrinogen level, viscoelastic monitoring, platelet count) are available. Conflicting opinions persist regarding the optimal initial strategy for coagulation support. Some experts, particularly in Europe, strongly advocate against empiric ratio-based transfusion without concurrent laboratory data guidance (goal-directed therapy) [388]. However, in the absence of rapid point-of-care coagulation testing to facilitate goal-directed therapy, initial treatment with blood components in a fixed ratio may be a reasonable pragmatic approach. If rapid coagulation results are available, they should be used to guide therapy.
In May 2005, an international expert conference on massive transfusion hosted by the US Army Institute of Surgical Research, drawing from experiences in the Iraq War, introduced the concept of immediate coagulation component administration in a 1:1:1 ratio of RBCs, plasma, and platelets [389–391]. This was proposed as a strategy for the initial resuscitation of massively bleeding patients until laboratory measurements could guide subsequent therapy. Subsequent retrospective evidence from military and civilian practice suggested improved outcomes with massive transfusion protocols incorporating early high-dose plasma therapy [392]. Numerous studies have since investigated this approach to determine if fixed-ratio administration of plasma and platelets relative to RBCs improves survival. Despite a substantial body of literature, the evidence regarding high-ratio transfusion remains conflicting. While many authors have suggested that early and aggressive plasma transfusion may reduce mortality [393], the optimal FFP:RBC and platelet:RBC ratios have been debated, complicated by potential survival bias in many observational studies [394, 395].
Survival bias arises from the fact that surviving patients are more likely to receive more plasma and platelets simply because they survive long enough to receive these blood products, skewing observational data. A prospective multicenter study involving a large cohort of patients undergoing massive transfusion suggested that high FFP:RBC and platelet:RBC ratios are associated with a survival benefit, even when accounting for time-dependency [225], while other studies have reached opposing conclusions [396]. Khan et al. found that FFP administration during the acute phase of ongoing bleeding did not consistently improve procoagulant factor levels or correct measures of clot function [396]. The recent Pragmatic, Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial, a large RCT involving 680 trauma patients suspected of or experiencing massive blood loss [397, 398], reported no significant difference in overall survival between early administration of plasma, platelets, and RBCs in a 1:1:1 ratio versus a 1:1:2 ratio. However, the 1:1:1 group did exhibit better anatomic hemostasis and fewer deaths due to exsanguination by 24 hours. Early use of platelets and higher FFP use in the 1:1:1 group was not associated with significantly increased complication rates. Early platelet administration, as detailed in Recommendation 29, is important, but practically, platelets may not always be readily available during the very initial resuscitation phase described here.
Complications associated with FFP transfusion, as with all human blood products, include circulatory overload, ABO incompatibility, infectious disease transmission (including prion diseases), and mild allergic reactions. Transfusion-related acute lung injury (TRALI) [399, 400], a severe complication, is linked to leukocyte antibodies in transfused plasma. The risk of TRALI has been significantly reduced by preferentially using plasma from male donors or women without a history of pregnancy [401]. Pathogen-inactivated plasma can further minimize infectious disease transmission risks.
Controversy also surrounds the use of plasma to address fibrinogen depletion in hemorrhagic shock. Fibrinogen is crucial for hemostasis, serving as a substrate for clot formation and a ligand for platelet aggregation. Fibrinogen is often the coagulation factor most significantly and earliest affected in trauma-induced coagulopathy. Many bleeding trauma patients with coagulopathy exhibit fibrinogen depletion below levels recommended for therapeutic supplementation. Schlimp et al. [159] showed that low fibrinogen levels (<1.5 g/l) are prevalent in patients with low admission Hb (<10 g/dl) or low base excess (<−6). Rourke et al. [402] found hypofibrinogenemia in 41% of hypotensive patients on admission. Hypotension, increasing shock severity, and higher injury severity scores (ISS ≥25) were all associated with reduced fibrinogen levels. Fibrinogen depletion is linked to poorer outcomes, and survival improves with fibrinogen administration [403]. Fibrinogen has the highest plasma concentration of all coagulation proteins. One liter of plasma contains approximately 2 grams of fibrinogen. Thus, for very early coagulation support, while awaiting viscoelastic or laboratory test results, administering 2 grams of fibrinogen can mimic the fibrinogen content of the first four RBC units in a 1:1 ratio and potentially correct pre-existing hypofibrinogenemia [385, 404]. Recent experimental data suggests that fibrinogen concentrate administration does not suppress endogenous fibrinogen synthesis [405].
Plasma transfusion, while potentially stabilizing fibrinogen levels, cannot effectively increase fibrinogen levels significantly unless very large volumes are infused [406]. The Activation of Coagulation and Inflammation in Trauma (ACIT) study [396] and a subsequent study by Khan et al. [15] confirmed that standard near 1:1 FFP:RBC transfusion protocols often fail to correct hypofibrinogenemia, and may even increase the proportion of coagulopathic patients over time. Only high-dose fibrinogen administration in the latter study resulted in improved coagulation and reduced coagulopathy. Furthermore, both FFP and pathogen-inactivated plasma require ABO group matching, thawing, and warming before administration. Unless pre-thawed plasma is available, plasma transfusion cannot be initiated as rapidly as universal RBC transfusion. Delays of approximately 90 minutes for thawed plasma have been reported [394, 407], potentially explaining why targeted plasma:RBC ratios are often achieved only hours after resuscitation initiation, during which time fibrinogen levels may remain suboptimal.
Antifibrinolytic Agents
Recommendation 25
We recommend that tranexamic acid be administered as early as possible to the trauma patient who is bleeding or at risk of significant hemorrhage at a loading dose of 1 g infused over 10 min, followed by an i.v. infusion of 1 g over 8 h. (Grade 1A)
We recommend that tranexamic acid be administered to the bleeding trauma patient within 3 h after injury. (Grade 1B)
We suggest that protocols for the management of bleeding patients consider administration of the first dose of tranexamic acid en route to the hospital. (Grade 2C)
Rationale
Tranexamic acid (TXA), a synthetic lysine analog, competitively inhibits plasminogen, an enzyme involved in fibrinolysis. TXA distributes widely throughout body tissues, with a plasma half-life of approximately 120 minutes [408]. The Clinical Randomisation of Antifibrinolytic therapy in Significant Haemorrhage (CRASH-2) trial [409], a landmark study, evaluated the effects of early TXA administration in trauma patients with or at risk of significant bleeding. This large RCT randomized over 20,000 adult trauma patients to receive either TXA (1 g loading dose over 10 min followed by 1 g infusion over 8 h) or placebo within 8 hours of injury. The primary outcome was in-hospital death within 4 weeks of injury.
The CRASH-2 trial demonstrated a statistically significant 1.5% reduction in all-cause mortality with TXA, and a 0.8% reduction in the risk of death specifically due to bleeding. TXA reduced bleeding-related deaths by approximately one-third, primarily by preventing exsanguination within the first 24 hours post-injury [410, 411]. One retrospective study has suggested that TXA may not be beneficial in patients with viscoelastic hyperfibrinolysis [412], while another found TXA to reduce multiple organ failure and mortality in severely injured shocked patients [413]. These discrepancies likely stem from methodological limitations in observational studies.
Concerns regarding potential thrombotic complications associated with lysine analogs like TXA and ε-aminocaproic acid have been raised. However, CRASH-2 demonstrated that TXA did not increase the rate of venous thromboembolism (VTE), and paradoxically, arterial thromboses, particularly myocardial infarction, were lower in the TXA group. No adverse events were directly attributed to TXA use in CRASH-2, although an increased seizure rate has been reported with high-dose TXA in cardiac surgery patients [414], potentially reflecting the role of fibrinolytic molecules as neurotransmitters.
An unplanned subgroup analysis of CRASH-2 data [415] revealed that early TXA administration (≤1 hour from injury) significantly reduced the risk of bleeding-related death by 2.5%. Treatment administered between 1 and 3 hours also showed a benefit, reducing bleeding-related death risk by 1.3%. However, TXA administration after 3 hours was associated with an increased risk of bleeding-related death by 1.3%. These findings strongly recommend that TXA should not be administered more than 3 hours post-injury. To facilitate early administration, protocols for managing bleeding trauma patients should consider incorporating pre-hospital TXA administration. Restricting TXA use to massive transfusion protocols or “high-risk” patients may capture only approximately 40% of the potential benefit [416]. To maximize benefit, TXA should be considered for all trauma patients with significant bleeding or risk of significant bleeding, integrated into standard trauma management protocols rather than solely reserved for massive hemorrhage situations.
Cost-effectiveness analyses of TXA use in trauma have been conducted in low-income (Tanzania), middle-income (India), and high-income (UK) countries [417, 418]. The cost of treating 1000 patients with TXA ranged from US$17,483 in Tanzania to US$30,830 in the UK. The estimated incremental cost per life-year gained was remarkably low, ranging from $48 to $66 across these diverse economic settings.
ε-aminocaproic acid, another synthetic lysine analog, is approximately tenfold less potent than TXA. It is administered with a loading dose of 150 mg/kg followed by a continuous infusion of 15 mg/kg/h. Its shorter elimination half-life (60–75 minutes) necessitates continuous infusion to maintain therapeutic levels until bleeding risk subsides. ε-aminocaproic acid may be considered as a viable alternative to TXA if TXA is unavailable.
Aprotinin, another antifibrinolytic agent, is not recommended for routine use in bleeding trauma patients due to safety concerns [419] and the demonstrated efficacy and safety of TXA.
Further Resuscitation
Goal-Directed Therapy
Recommendation 26
We recommend that resuscitation measures be continued using a goal-directed strategy guided by standard laboratory coagulation values and/or viscoelastic tests. (Grade 1C)
Rationale
Treatment of bleeding trauma patients is often guided by the principle that normalizing coagulation parameters will improve patient outcomes, although direct evidence definitively supporting this assumption is limited. During initial resuscitation, the patient’s coagulation status is unknown until laboratory results are available. Therefore, initial blood product and therapeutic interventions are often based on a “best guess” approach, with local protocols varying due to a lack of definitive evidence for a single optimal “formula.” These “best guess” protocols typically involve pre-defined ratios of RBCs, FFP, and other hemostatic components administered in “bundles” or “packs.”
During subsequent resuscitation phases, as more information from laboratory or point-of-care coagulation tests becomes available, treatment strategies should be refined and transitioned to a goal-directed approach. If initial laboratory data are unavailable, initiating “best guess” treatment based on the presumption of coagulopathy in severely injured patients is reasonable. However, further resuscitation efforts should be guided by a goal-directed strategy.
Clinicians must be mindful of the inherent time lag between sample collection and result availability. Treatment decisions should not be delayed solely to await test results. Delays in coagulation results pose a greater challenge in settings lacking point-of-care testing capabilities. To mitigate the risk of lagging behind dynamic changes in patient condition, treatment decisions should integrate both available test results and the clinician’s informed judgment regarding potential shifts in the patient’s coagulation status since the sample was drawn. Specific therapeutic goals for goal-directed therapy are further explored in the subsequent sections.
Fresh Frozen Plasma
Recommendation 27
If a plasma-based coagulation resuscitation strategy is used, we recommend that plasma (FFP or pathogen-inactivated plasma) be administered to maintain PT and APTT <1.5 times normal. (Grade 1C)
We recommend that plasma transfusion be avoided in patients without substantial bleeding. (Grade 1B)
Rationale
Fresh frozen plasma (FFP) and pathogen-inactivated plasma have been widely used for decades as a source of coagulation factors in bleeding patients globally. FFP typically contains approximately 70% of normal levels of all clotting factors, making it a seemingly logical replacement fluid. However, significant variability exists in the coagulation factor content across different FFP preparations [256]. We recommend FFP use within a plasma-based coagulation resuscitation strategy when evidence of coagulation factor deficiency is present, as indicated by prolonged PT and APTT exceeding 1.5 times the normal control range, or by viscoelastic testing abnormalities. While dedicated RCTs evaluating this specific strategy are lacking, this approach is clinically widespread. Careful monitoring of hemorrhage management is essential to ensure appropriate FFP transfusion, as FFP administration carries potential risks, including circulatory overload, allergic reactions, and TRALI.
Prolongation of “clotting time” or “reaction time” on viscoelastic testing may also be considered an indication for FFP administration. However, robust scientific evidence supporting this specific indication is limited, and normalization of fibrinogen levels (as per Recommendation 28) often improves these viscoelastic parameters.
Fibrinogen and Cryoprecipitate
Recommendation 28
If a concentrate-based strategy is used, we recommend treatment with fibrinogen concentrate or cryoprecipitate if significant bleeding is accompanied by viscoelastic signs of a functional fibrinogen deficit or a plasma fibrinogen level of less than 1.5–2.0 g/l. (Grade 1C)
We suggest an initial fibrinogen supplementation of 3–4 g. This is equivalent to 15–20 single donor units of cryoprecipitate or 3–4 g fibrinogen concentrate. Repeat doses must be guided by viscoelastic monitoring and laboratory assessment of fibrinogen levels. (Grade 2C)
Rationale
Fibrinogen is a critical component of the coagulation cascade, serving as the final substrate for clot formation and a key ligand for platelet aggregation [280, 420]. Hypofibrinogenemia is a frequent and early feature of complex coagulopathies associated with massive bleeding. Fibrinogen levels often decrease rapidly in severely injured trauma patients, and low fibrinogen levels are associated with increased transfusion requirements and higher mortality [421]. Given that the body has no readily accessible fibrinogen reserves beyond plasma, the total fibrinogen pool in an adult is relatively small (approximately 10g in an 80kg individual), making rapid compensation for fibrinogen depletion challenging. Schlimp et al. [159] demonstrated a strong correlation between admission fibrinogen levels and readily obtainable routine laboratory parameters like Hb and base excess. Hypofibrinogenemia (<1.5 g/l) was prevalent in trauma patients with low admission Hb (<10 g/dl) or low base excess (<−6). Rourke et al. [402] observed hypofibrinogenemia in 41% of patients hypotensive on admission, with hypotension, increasing shock severity, and high injury severity (ISS ≥25) all associated with reduced fibrinogen levels.
Fibrinogen depletion is consistently associated with poorer outcomes, and survival tends to improve with fibrinogen administration [403]. In coagulopathic civilian trauma patients, median fibrinogen concentrations are often significantly reduced (e.g., 0.9 g/l, IQR 0.5–1.5 g/l) in conjunction with impaired clot firmness on viscoelastic testing [191]. Healthy volunteers typically exhibit significantly higher fibrinogen levels and clot firmness [191]. In trauma patients, a thromboelastometry maximum clot firmness (MCF) of 7 mm was associated with a fibrinogen level of approximately 1.5–2.0 g/l [191]. During postpartum hemorrhage, fibrinogen plasma concentration is the only coagulation parameter independently associated with progression towards severe bleeding, with levels <2 g/l predictive of progression to severe hemorrhage [422].
An early observational study suggested that fibrinogen supplementation may improve survival in combat-related trauma [403]. In civilian trauma settings, thromboelastometry-guided fibrinogen replacement has been shown to reduce allogeneic blood product exposure [12, 378, 385]. Retrospective single-center studies have also suggested reduced morbidity and mortality rates in select patient populations with damage control resuscitation strategies that include fibrinogen replacement [320, 378, 423]. However, robust, adequately powered prospective clinical trials definitively demonstrating the risk-benefit profile of fibrinogen supplementation in bleeding trauma patients are still lacking [424, 425]. It has been suggested that fibrinogen dosage may be estimated based on thromboelastometric monitoring. For example, administering 0.5 g of fibrinogen to an 80 kg patient may increase the A10 MCF (amplitude at 10 minutes) by approximately 1 mm, potentially facilitating predictable and rapid fibrinogen level increases to a target level [426].
The retrospective Military Application of Tranexamic Acid in Trauma Emergency Resuscitation (MATTERs II) study of massive military bleeding suggested that cryoprecipitate may independently enhance the survival benefit of TXA in severely injured patients requiring transfusion [427]. However, cryoprecipitate administration is often delayed in practice. In the Prospective, Observational, Multicenter, Major Trauma Transfusion (PROMMTT) study [428], the median time from admission to the first cryoprecipitate unit was nearly 3 hours. In the ACIT study [396], cryoprecipitate was administered only after the first six units of blood. A small feasibility RCT suggested that early cryoprecipitate administration in trauma patients is achievable [429].
Methodological challenges exist in measuring fibrinogen concentration accurately [430, 431]. The Clauss method is commonly recommended, but may overestimate fibrinogen levels in the presence of artificial colloids like HES, although it remains a gold standard due to its direct measurement of fibrinogen function [431]. Fibrinogen thromboelastometry is also influenced by Hct [432] and factor XIII levels [433].
The question of whether fibrinogen administration, via concentrate, cryoprecipitate, or FFP, increases the risk of hospital-acquired VTE has not been systematically addressed. However, fibrinogen levels naturally increase as part of the acute phase response after major surgery and trauma [371, 434–436], even without intraoperative fibrinogen administration. Interestingly, intraoperative fibrinogen concentrate administration in trauma patients [371] or cardiac surgery patients resulted in elevated intra- and early postoperative fibrinogen levels, but fibrinogen levels were comparable on postoperative days 1–7 in patients with and without intraoperative fibrinogen administration [436, 437].
The rationale for fibrinogen supplementation should be considered in conjunction with that for plasma transfusion (Recommendation 27). Insufficient evidence currently exists to definitively recommend one strategy over the other, or whether a combined approach using both strategies might be optimal.
Platelets
Recommendation 29
We recommend that platelets be administered to maintain a platelet count above 50 × 10^9/l. (Grade 1C)
We suggest maintenance of a platelet count above 100 × 10^9/l in patients with ongoing bleeding and/or TBI. (Grade 2C)
If administered, we suggest an initial dose of four to eight single platelet units or one aphaeresis pack. (Grade 2C)
Rationale
Platelets are crucial for hemostasis after injury, but the optimal role of platelet transfusion in trauma remains debated. Historically, platelet transfusion thresholds were based on critical platelet counts. One small prospective study in massively transfused patients identified a platelet count of 9 x 10^9/l as a threshold for diffuse bleeding [438], while another suggested a threshold of 20 x 10^9/l [439]. However, an older prospective randomized trial evaluating prophylactic platelet transfusion in massively transfused patients concluded that platelet administration did not reduce microvascular nonsurgical bleeding [440]. More recently, it has been shown that low or decreasing platelet counts in trauma patients are predictive of higher mortality [441], and proactive platelet administration in massively bleeding patients with ruptured abdominal aortic aneurysms improved survival when platelet counts were maintained above 50 x 10^9/l, with further improvements above 100 x 10^9/l [442].
Low platelet counts also predict intracranial hemorrhage (ICH) progression and mortality after TBI [443, 444]. In blunt TBI patients, a platelet count ≤100 x 10^9/l was an independent predictor of ICH progression, need for neurosurgical intervention, and mortality [445]. However, platelet transfusion did not improve outcomes in TBI patients with moderate thrombocytopenia (50–107 x 10^9/l) [446]. Currently, limited high-quality evidence supports specific platelet count thresholds for platelet transfusion in trauma patients.
The standard therapeutic platelet dose is one concentrate (60–80 x 10^9 platelets) per 10 kg body weight. One apheresis platelet product, roughly equivalent to six whole blood-derived units, contains approximately 3–4 x 10^11 platelets in 150–450 ml of donor plasma [447, 448], depending on collection practices. Platelet-rich plasma used in the United States typically contains fewer platelets than high-output platelet concentrates collected by apheresis or pooling buffy coats, primarily used in Europe [449]. This difference should be considered when interpreting studies supporting higher platelet transfusion ratios. A dose of four to eight platelet units or one apheresis unit is generally sufficient to achieve hemostasis in thrombocytopenic bleeding patients and should increase platelet counts by 30–50 x 10^9/l [375]. However, the usual 60–70% recovery rate in peripheral blood may be lower in conditions with increased platelet consumption [449]. ABO-identical or ABO-compatible platelet transfusions are recommended for optimal platelet yield [448].
Early, preemptive platelet administration in non-thrombocytopenic massively bleeding patients remains controversial. In acute blood loss, bone marrow and spleen release platelets, delaying a decrease in peripheral blood platelet counts. Consequently, platelet counts are often within the normal range (150–400 x 10^9/l) during early traumatic coagulopathy [441, 450–452]. Platelet counts <100 x 10^9/l on admission are relatively uncommon, observed in <5% to ~18% of trauma patients [450, 452]. In a large cohort study, platelet counts decreased significantly within 2 hours of hospital admission and continued to decline over the subsequent 22 hours, suggesting a dynamic interplay between injury, resuscitation, and platelet consumption [441]. A platelet count of 50 x 10^9/l may be anticipated when approximately two blood volumes have been replaced with fluids or red cell components [421].
Admission platelet count may have prognostic value. Lower platelet counts on admission have been associated with increased injury severity [450], morbidity [443], and mortality [450, 451, 453] in massively transfused trauma patients. This association extends even within the normal platelet count range [441, 451], suggesting that a “normal” platelet count may be insufficient for effective cellular hemostasis after severe trauma. Platelet count alone, however, is a limited indicator of transfusion need, as it does not account for platelet function.
Emerging evidence underscores the critical role of platelet dysfunction in traumatic coagulopathy pathophysiology [454, 455]. Moderate or even mild reductions in platelet aggregation are strongly associated with increased mortality [214, 456, 457]. Early platelet dysfunction, even before significant fluid or blood product administration, is observed in trauma patients and persists during resuscitation, suggesting a potential role for early platelet transfusion in managing traumatic coagulopathy [455]. In TBI patients, platelet transfusion has been shown to reverse aspirin-like platelet inhibition in some cases [458], but not collagen-induced dysfunction [459].
While retrospective and observational studies suggest potential benefits of higher platelet:RBC ratios in massively transfused trauma patients [460–466], these findings are susceptible to biases and confounding factors. A meta-analysis [467] and systematic review [468] investigating platelet transfusion impact in severe trauma and massive transfusion showed improved survival with high platelet:RBC ratios, but the evidence base remains primarily observational. The recent Pragmatic, Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial, a large RCT, found no difference in overall survival with 1:1:1 versus 1:1:2 transfusion ratios, but the 1:1:1 group exhibited better hemostasis and fewer deaths due to exsanguination by 24 hours [397]. However, the PROPPR trial did not isolate and evaluate the independent effects of platelet transfusion.
Theoretical concerns exist regarding potential over-transfusion of plasma and platelets in ratio-driven resuscitation, potentially leading to no benefit or increased morbidity like multiple organ failure [466, 480]. Recent data suggest that both early FFP (0–6 hours) and delayed platelet (7–24 hours) transfusions may be risk factors for hypoxemia and ARDS after 24 hours, respectively [481]. The age of transfused platelets may also influence outcomes [482]. While some studies report reduced morbidity with aggressive plasma and platelet use [382, 463, 464], robust evidence supporting routine early prophylactic platelet transfusion within massive transfusion protocols remains limited [483].
Calcium
Recommendation 30
We recommend that ionized calcium levels be monitored and maintained within the normal range during massive transfusion. (Grade 1C)
Rationale
Acute hypocalcemia is a common complication of massive transfusion [484]. Citrate anticoagulant in stored blood binds calcium, potentially reducing serum ionized calcium levels [485]. Two observational cohort studies demonstrated that low admission ionized calcium levels are associated with increased mortality and a higher need for massive transfusion [486, 487]. In one study, hypocalcemia within the first 24 hours was a better predictor of mortality and massive transfusion need than fibrinogen levels, acidosis, or platelet counts [486]. Admission ionized calcium measurement may facilitate early identification of patients likely to require massive transfusion, enabling timely blood product preparation and administration. However, direct evidence demonstrating that preventing ionized hypocalcemia improves mortality in massively transfused patients is lacking.
Extracellular calcium exists in a free ionized state (45%) and protein-bound, inactive state (55%). Normal ionized calcium concentration ranges from 1.1 to 1.3 mmol/l and is influenced by pH (a 0.1 pH unit increase decreases ionized calcium by ~0.05 mmol/l) [488]. Ionized calcium availability is essential for fibrin polymerization and stabilization, and reduced cytosolic calcium impairs platelet function [489]. Low ionized calcium also compromises cardiac contractility and systemic vascular resistance. Maintaining ionized calcium within the normal range is therefore crucial for both coagulation and cardiovascular function [489].
Early hypocalcemia following trauma is correlated with FFP and colloid transfusion volumes, but not crystalloid volume. Hypocalcemia is more common with FFP and platelet transfusion due to their citrate content. Citrate is rapidly metabolized by the liver, and hypocalcemia is usually transient during standard transfusion. However, citrate metabolism can be severely impaired by hypoperfusion, hypothermia, and hepatic insufficiency [489].
Antiplatelet Agents
Recommendation 31
We suggest administration of platelets in patients with substantial bleeding or intracranial hemorrhage who have been treated with antiplatelet agents. (Grade 2C)
We suggest the measurement of platelet function in patients treated or suspected of being treated with antiplatelet agents. (Grade 2C)
We suggest treatment with platelet concentrates if platelet dysfunction is documented in a patient with continued microvascular bleeding. (Grade 2C)
Rationale
Conflicting data exist regarding the effects of antiplatelet agents (APAs), primarily aspirin and clopidogrel, on traumatic bleeding. Studies of non-elective orthopedic procedures report either increased perioperative blood loss [490, 491] or no effect [492–494] in patients on preoperative APAs. The need for blood transfusion in orthopedic patients on APAs is also debated, with some studies reporting higher transfusion rates [491, 495, 496] and others reporting no difference compared to controls [492–494, 497, 498]. Pre-injury APA use did not affect morbidity or mortality in retrospective studies of patients with pelvic fractures [495] or general trauma without brain injury [499], but showed conflicting effects on early hip fracture surgery outcomes [491, 494, 497, 498, 500]. One observational study found aspirin use associated with increased postoperative transfusion needs and mortality after hip fracture surgery [491]. However, retrospective studies on hip fracture surgery in patients on clopidogrel reported similar postoperative outcomes compared to APA-naïve patients [497, 498, 500, 501], except for longer hospital stays in some studies [494, 498].
The role of pre-injury APAs in traumatic intracranial hemorrhage (ICH) development is also controversial [502–506]. One observational study reported a fivefold increased risk of traumatic ICH in patients on APAs [502]. Even minor head trauma (GCS 14–15) in APA users was associated with a higher ICH incidence [507–509], necessitating longer observation periods for delayed ICH [510, 511]. Other studies failed to demonstrate this association [503, 504, 506], although one reported clopidogrel use significantly associated with ICH after minor trauma [512].
The relationship between outcome and pre-injury APA in ICH is also debated [504, 508, 513–518], as is the relationship between APA and outcome in stroke literature [519–522]. One meta-analysis found pre-ICH APA use associated with modestly increased mortality but little impact on functional outcomes [523]. Another meta-analysis of blunt head trauma patients found only a slight, non-significant increase in mortality in pre-injury APA users [524]. However, more recent studies report conflicting findings, with some showing associations between APA use and worsened ICH [525, 526] or increased neurosurgical intervention [526], while others report no influence on survival or neurosurgical intervention [527].
Limited data specifically address the impact of individual APAs. Studies analyzing clopidogrel use prior to both spontaneous and traumatic ICH reported worsened outcomes compared to controls, including increased mortality [518, 520], morbidity [528], lesion progression [503, 508, 520, 529], and need for long-term care facilities [518, 520]. Pre-injury aspirin did not affect outcomes in mild-to-moderate head injury [504, 508] or mortality [458] in observational studies, but one RCT reported increased hemorrhage volume and mortality [531]. Surprisingly, reduced platelet activity has been observed in ICH patients without known aspirin use [458, 532] and was associated with greater ICH volume growth and poorer 3-month outcomes [533]. Greater platelet inhibition was found in patients on dual APA therapy compared to single-agent APA [532].
Lower platelet counts also pose additional risks. TBI patients on pre-hospital APAs with platelet counts ≤135 x 10^9/l were significantly more likely to experience ICH progression, and those with platelet counts ≤95 x 10^9/l were significantly more likely to require neurosurgical intervention [444].
These findings, coupled with the 20-30% rate of non-responsiveness to aspirin and/or clopidogrel [535], suggest that reliable platelet function measures would be valuable in bleeding trauma patients to guide platelet transfusion or reversal agent use. Patients with occult platelet dysfunction who would benefit from platelet transfusion could be identified [536], and unnecessary platelet transfusions avoided [458].
Currently, no consensus exists on the optimal platelet function assay, and the role of platelet transfusion in APA-associated ICH is debated. Guidelines for ICH management in APA users suggest platelet transfusion with a low-grade recommendation [537], and it is indicated for clopidogrel-associated traumatic hemorrhage, although its clinical utility remains to be fully established [538]. Retrospective studies have not shown outcome benefits from platelet transfusion in APA users with spontaneous [521, 522, 539] or traumatic [514, 540, 541] ICH. A meta-analysis of small studies on platelet transfusion in APA-associated ICH found no clear benefit [542]. However, the timing of platelet transfusion may have been suboptimal in some studies [533, 539], and a small prospective study suggested that early platelet transfusion may improve platelet activity and outcomes in non-traumatic ICH [543].
In vitro studies in healthy volunteers on aspirin and clopidogrel suggest that two to three platelet pools may normalize platelet function [544]. However, platelet transfusion efficacy in reversing platelet dysfunction in traumatic ICH remains conflicting and inconclusive [458, 459, 545–547]. Platelet transfusion improved aspirin-induced platelet dysfunction in some studies [458, 545], but not in others [546], and not consistently in clopidogrel-treated patients [545]. One study showed improved platelet aggregation with ex vivo platelet supplementation, regardless of APA use [547]. However, while aspirin effects were fully reversed, ADP-dependent aggregation recovery was limited even with high platelet doses [547]. A small prospective trial showed that platelet transfusion improved aspirin-induced, but not collagen-induced, platelet dysfunction in TBI patients [459]. Divergent results may be due to varying platelet doses transfused, from one pack [546] to multiple apheresis units [458]. Another potential explanation for limited platelet transfusion benefit is that ongoing APA effects may inactivate transfused platelets [543]. Results from a multicenter RCT on platelet transfusion in APA-associated ICH are pending [548].
Suggested doses to normalize platelet activity in healthy volunteers on APAs range from five platelet units for aspirin alone to ten to fifteen units for aspirin and clopidogrel [544]. Successful perioperative management of patients on aspirin and clopidogrel undergoing urgent surgery using two apheresis platelet units has been reported [549]. Given that clopidogrel’s active metabolite persists and transfused platelets have a short half-life, repeat platelet transfusions may be warranted [550].
Beyond platelet transfusion, potential APA reversal strategies include desmopressin and recombinant activated coagulation factor VII (rFVIIa) [538]. The rationale for desmopressin in aspirin-treated patients is discussed in Recommendation 32. In healthy volunteers, rFVIIa reversed the inhibitory effects of aspirin and clopidogrel [551], with effective doses lower than those used in hemophilia patients [552]. Tranexamic acid (TXA) has also been shown to partially improve platelet function in dual antiplatelet therapy patients [553]. Fibrinogen concentrate has also demonstrated potential effectiveness in improving hemostasis in trauma patients on APAs [554].
Desmopressin
Recommendation 32
We suggest that desmopressin (0.3 μg/kg) be administered in patients treated with platelet-inhibiting drugs or with von Willebrand disease. (Grade 2C)
We do not suggest that desmopressin be used routinely in the bleeding trauma patient. (Grade 2C)
Rationale
Desmopressin (DDAVP; 1-deamino-8-D-arginine vasopressin) enhances platelet adherence and aggregate formation on artery subendothelia and is the primary treatment for bleeding in von Willebrand disease, a relatively common disorder (prevalence ~1 in 100) [555, 556]. Meta-analyses in patients without von Willebrand disease [557, 558] suggest desmopressin may reduce perioperative blood loss [557] or modestly reduce transfusion needs [558]. Patients with impaired platelet function, as assessed by platelet function analyzers [559] or whole blood multiple electrode aggregometry [560], may benefit from desmopressin. Thrombotic complications, initially a concern [561], were not confirmed in a recent meta-analysis [558].
Desmopressin improves platelet function in volunteers on aspirin [562] and clopidogrel [563], and perioperatively in patients with mild inherited platelet defects [564]. One older meta-analysis suggested desmopressin benefits in aspirin users [565], and desmopressin has been recommended for platelet inhibitor-associated ICH [538, 566]. The standard dose is 0.3 μg/kg diluted in 50 ml saline infused over 30 minutes [564]. Recent prospective studies have shown desmopressin improves platelet function in ICH patients with or without aspirin use [567, 568]. Platelet function assessment using PFA-100 [559] or whole blood MEA [560] may help identify patients who could benefit from desmopressin. Combined platelet concentrate and desmopressin administration has been suggested to enhance platelet function recovery [569], but a recent study found no benefit of desmopressin and platelet transfusion in early ICH progression or mortality in TBI patients [570].
Desmopressin appears effective in mitigating platelet inhibition by ADP receptor inhibitors like clopidogrel [571] and ticagrelor [572], but data for prasugrel are limited. Desmopressin has been recommended in patients on platelet inhibitors with intracerebral bleeding [538, 566] and in trauma patients with von Willebrand disease [573]. Desmopressin also prevents hypothermia-induced primary hemostasis impairment [574] and increases platelet aggregation under hypothermia and acidosis conditions [575].
Limited studies exist on desmopressin use in general trauma, ICH, or TBI [538]. However, desmopressin improved platelet function and von Willebrand factor levels in ICH patients with reduced platelet activity and/or aspirin use [568]. Conversely, co-administration of desmopressin with platelet transfusion did not slow early ICH progression in a recent retrospective TBI study [570]. Nevertheless, desmopressin is recommended in platelet inhibitor-treated patients with intracerebral bleeding [538, 566] and in trauma patients with von Willebrand disease [573]. Desmopressin’s ability to counteract hypothermia and acidosis-induced platelet dysfunction further supports its potential role in trauma settings [574, 575].
Prothrombin Complex Concentrate
Recommendation 33
We recommend the early use of prothrombin complex concentrate (PCC) for the emergency reversal of vitamin K-dependent oral anticoagulants. (Grade 1A)
We suggest the administration of PCC to mitigate life-threatening post-traumatic bleeding in patients treated with novel oral anticoagulants. (Grade 2C)
Provided that fibrinogen levels are normal, we suggest that PCC or plasma be administered in the bleeding patient based on evidence of delayed coagulation initiation using viscoelastic monitoring. (Grade 2C)
Rationale
Prothrombin complex concentrate (PCC) is superior to FFP for rapid vitamin K antagonist reversal [576–578], with evidence of reduced hematoma formation in TBI patients [579, 580]. PCC is therefore the preferred agent for reversing vitamin K antagonist effects [581].
No universally accepted reversal strategies exist for non-vitamin K antagonist oral anticoagulants (NOACs), but despite limited clinical evidence, PCC has been used anecdotally to reverse NOAC effects [582–586]. Animal studies and ex vivo human studies support PCC/aPCC and recombinant factor VIIa in reversing NOACs [582]. PCC’s ability to normalize coagulation tests, viscoelastic parameters, and thrombin generation depends on dosage [611–619]. Whether this translates to improved hemostasis and reduced bleeding may depend on NOAC plasma levels. No effect on bleeding was seen at high rivaroxaban concentrations in rabbits [609], while reduced bleeding was found at lower dabigatran concentrations in mice [620]. In rats, increasing PCC doses reversed bleeding volume [621], but high NOAC plasma levels may require very high PCC doses for effective reversal [621].
Measuring plasma NOAC concentrations is recommended to assess their potential impact on coagulation [622]. Anti-factor Xa activity tests can quantify factor Xa antagonists (rivaroxaban, apixaban, edoxaban). If unavailable, LMWH anti-factor Xa activity assays can provide qualitative information. Dabigatran-calibrated thrombin time can quantify thrombin inhibitors. Thrombin time and APTT can qualitatively detect dabigatran presence if specific assays are unavailable. Viscoelastic tests are affected by factor Xa and IIa inhibitors [623], but they provide an overall coagulation snapshot and cannot specifically quantify NOAC effects.
If NOAC treatment is known or suspected, and anti-factor Xa activity is detected, high-dose (25–50 U/kg) PCC/aPCC may be initiated. We suggest starting with 25 U/kg, repeated as needed due to PCC/aPCC thrombotic potential [599]. For dabigatran (anti-FIIa activity), idarucizumab (5 g IV), the specific dabigatran antidote [624, 625], should be used if available. If idarucizumab is unavailable, preoperative hemodialysis may be considered [626, 627]. Concomitant TXA administration is generally recommended in trauma patients (Recommendation 25). Recombinant factor VIIa use has been described but is not recommended as first-line treatment. Hematologist consultation is advisable for NOAC reversal management.
As of late 2015, idarucizumab, the dabigatran antidote, had received FDA and EMA approval. Specific antidotes for factor Xa inhibitors, such as andexanet alfa [629], are under development but not yet clinically approved [630, 631].
Recombinant Activated Coagulation Factor VII
Recommendation 36
We suggest that the off-label use of rFVIIa be considered only if major bleeding and traumatic coagulopathy persist despite all other attempts to control bleeding and best-practice use of conventional hemostatic measures. (Grade 2C)
Rationale
Recombinant activated factor VIIa (rFVIIa) should be reserved as a last-resort adjunctive therapy, considered only when major bleeding and traumatic coagulopathy persist despite exhausting all other bleeding control measures and optimal conventional hemostatic management. This includes surgical interventions, best-practice blood product administration (RBCs, platelets, FFP, cryoprecipitate/fibrinogen, targeting Hct >24%, platelets >50 x 10^9/l, and fibrinogen >1.5–2.0 g/l), antifibrinolytics, and correction of severe acidosis, hypothermia, and hypocalcemia.
rFVIIa relies on the patient’s intrinsic coagulation system, necessitating adequate platelet counts and fibrinogen levels to support its activity [632, 633]. Physiological pH and body temperature are also crucial, as even mild acidosis and hypothermia impair coagulation enzyme kinetics [299, 300, 634]. Predictors of poor rFVIIa response include pH <7.2 [635, 636], temperature <34°C [635, 637], platelet count <50 x 10^9/l [635], fibrinogen <1 g/l [635], and blood pressure ≤90 mmHg [635]. One study reported no survival benefit with rFVIIa administration when pH was <7.10 [636]. Another study from the Australian and New Zealand Haemostasis Registry found rFVIIa ineffective in patients with a median pH of 7.1 [637]. Hypocalcemia, frequently present in severely injured patients [638], also impairs coagulation. Ionized calcium monitoring and intravenous calcium administration may be necessary [639].
Despite numerous case studies and series suggesting rFVIIa’s benefit in trauma bleeding, high-quality evidence is limited [640–643]. A multicenter, randomized, placebo-controlled trial in blunt and penetrating trauma patients [644] found that rFVIIa (200 μg/kg initial dose followed by 100 μg/kg doses at 1 and 3 hours) reduced RBC transfusion requirements and massive transfusion need in blunt trauma patients surviving >48 hours, also reducing ARDS incidence. However, no significant effects were seen in penetrating trauma patients, although trends toward reduced transfusion needs were observed. Other retrospective studies and case reports have shown similar trends [645–647]. Another RCT (CONTROL trial) [648], evaluating rFVIIa as adjunct therapy in major trauma patients bleeding despite damage control resuscitation and surgery, was terminated early due to slow enrollment and low mortality, precluding definitive conclusions. Thrombotic adverse events were similar between rFVIIa and placebo groups in CONTROL trial.
A German trauma registry study comparing matched groups with and without early rFVIIa administration found no difference in mortality or transfusion needs, but an increased multiple organ failure incidence in the rFVIIa group [649]. In a retrospective thromboelastographic-guided hemostatic therapy study in abdominal trauma patients, rFVIIa administration was associated with improved R time and reduced blood product transfusion compared to controls [650].
Conversely, rFVIIa use in isolated head injury may be harmful. A case-controlled study in traumatic ICH patients suggested increased death risk with rFVIIa, regardless of injury severity [651]. Cochrane systematic reviews have found no reliable evidence from RCTs supporting hemostatic drug effectiveness in reducing mortality or disability in TBI patients [652]. rFVIIa did not improve mortality or reduce plasma use in warfarin-treated TBI patients [653]. Given the lack of evidence supporting rFVIIa use in isolated head trauma, the previous negative recommendation against rFVIIa use in isolated head trauma has been removed from this version of the guideline as self-evident.
The optimal rFVIIa dose remains debated. Dosing used in RCTs was based on European expert recommendations [654], while Israeli guidelines, based on a compassionate-use case series [641], proposed higher initial doses (100–140 μg/kg). Pharmacokinetic modeling suggests the RCT dosing regimen can achieve adequate plasma drug levels for hemostasis [655]. Bain et al. compared a lower-dose rFVIIa protocol to prior higher-dose practice and found no outcome differences despite lower total rFVIIa doses in the post-protocol group [656].
A recent prospective non-randomized trial in TBI patients with coagulopathy found rFVIIa (20 μg/kg) improved INR but not hospital mortality [657].
If rFVIIa is considered, informed consent regarding off-label use and potential thromboembolic risks [658] should be obtained. Meta-analyses of rFVIIa use outside approved indications have shown a higher risk of arterial thromboembolic events [659], although trauma studies have not consistently demonstrated increased thromboembolic risk [660]. A recent retrospective cohort study in surgical and trauma patients receiving off-label rFVIIa reported a 12.5% thromboembolic event incidence, with higher incidence in cardiothoracic surgery and penetrating trauma, but no dose-dependent increase [661]. Elevated endogenous thrombin potential following PCC administration in trauma patients, not reflected in standard coagulation tests, raises further caution regarding pro-thrombotic risks with factor concentrate use [371]. Therefore, early thromboprophylaxis after bleeding control is crucial in patients receiving PCC or rFVIIa.
Thromboprophylaxis
Recommendation 37
We recommend pharmacological thromboprophylaxis within 24 h after bleeding has been controlled. (Grade 1B)
We recommend early mechanical thromboprophylaxis with intermittent pneumatic compression (IPC) (Grade 1C) and suggest early mechanical thromboprophylaxis with anti-embolic stockings. (Grade 2C)
We do not recommend the routine use of inferior vena cava filters as thromboprophylaxis. (Grade 1C)
Rationale
The risk of hospital-acquired venous thromboembolism (VTE) after multiple trauma is high, exceeding 50%, with pulmonary embolism (PE) being the third leading cause of death in patients surviving beyond the initial 3 days [662]. Limited RCTs have investigated thromboprophylaxis in trauma patients, and anti-embolic stocking efficacy has not been specifically evaluated in this population. A meta-analysis found IPC ineffective in reducing DVT rates [663]; however, mechanical methods are widely used due to their low bleeding risk profile.
A systematic review and meta-analysis [664] showed that heparin thromboprophylaxis (any type) reduces DVT and PE in medical-surgical critically ill patients. Low molecular weight heparin (LMWH) compared to twice-daily unfractionated heparin (UFH) reduced both overall and symptomatic PE rates. Heparin thromboprophylaxis did not significantly impact major bleeding or mortality rates in the ICU setting. A study of VTE development in 289 critically ill patients found thromboprophylaxis failure more likely with elevated BMI, VTE history, and vasopressor use [665].
Heparin-induced thrombocytopenic thrombosis is a known heparin complication, more frequent with UFH than LMWH. Trauma severity correlates with heparin-induced thrombocytopenia risk, making platelet count monitoring essential in trauma patients receiving heparin [666]. In summary, heparin use after hemostasis is the most effective thromboprophylaxis option in trauma. Mechanical methods are preferred in patients with bleeding risk. Due to varied trial results comparing UFH and LMWH, neither is definitively recommended over the other. LMWH, primarily renally excreted, carries a risk of accumulation in renal failure patients, requiring dose adjustments and/or monitoring.
Contraindications to pharmacological thromboprophylaxis include patients already on full-dose anticoagulation, significant thrombocytopenia (platelet count <50 x 10^9/l), untreated bleeding disorders, active bleeding, uncontrolled hypertension (>230/120 mmHg), planned lumbar puncture/spinal analgesia within 12 hours or recent lumbar puncture/spinal analgesia (within 4 hours, or 24 hours if traumatic), high bleeding risk procedures, or new hemorrhagic stroke. However, a systematic review found pharmacological thromboprophylaxis appears safe in TBI patients with stabilized hemorrhage patterns [667].
Prophylactic inferior vena cava filters are commonly used, but no evidence supports added benefit over pharmacological thromboprophylaxis. PE can still occur despite filters, and filters have both short- and long-term complication risks, high costs, and may create a false sense of security, delaying effective pharmacological prophylaxis. Furthermore, IVC filters require a second invasive procedure for removal.
Optimal pharmacological thromboprophylaxis timing is challenging. Data from 175,000 critical care admissions showed higher mortality in patients not receiving thromboprophylaxis within the first 24 hours [668], reflecting that bleeding patients may have a higher VTE risk than non-bleeding patients [669].
Limited research exists on mechanical thromboprophylaxis in critical care. The recent CLOTS 3 study, a large RCT in stroke patients, demonstrated IPC benefit, reducing DVT rates by 3.6% and showing a non-significant mortality reduction [670]. While stroke patients differ from critical care patients, both share risk factors (immobility, acute-phase response), leading to an upgraded recommendation for IPC. Anti-embolic stockings are suggested as another mechanical prophylaxis option, although evidence in trauma patients remains limited.
Guideline Implementation and Quality Control
Guideline Implementation
Recommendation 38
We recommend the local implementation of evidence-based guidelines for management of the bleeding trauma patient. (Grade 1B)
Assessment of Bleeding Control and Outcome
Recommendation 39
We recommend that local clinical quality and safety management systems include parameters to assess key measures of bleeding control and outcome. (Grade 1C)
Rationale
Evidence supporting patient management algorithm effectiveness in changing clinical care is limited, but local implementation of multidisciplinary, evidence-based guidelines or algorithms for bleeding trauma patients is likely to enhance awareness among specialties and improve interdisciplinary understanding. Local algorithms, within the framework of available evidence, allow flexibility to accommodate local pre-hospital conditions, diagnostic and therapeutic resources, and improve care consistency. However, guidelines are designed for the “average” patient, requiring clinicians to adapt and tailor treatment to individual cases.
Implementing key guideline interventions is likely to improve outcomes [671, 672] and reduce mortality and complications [673]. Furthermore, guideline-concordant treatment may be cost-saving [674]. However, strict guideline adherence can be challenging in complex, poor-prognosis cases, making the association between guideline adherence and positive outcomes not necessarily causal.
Checklist-based implementation, similar to the Safe Surgery Initiative [675], may facilitate guideline adoption, potentially reducing postoperative complications [676]. Alternatively, bundle implementation, as successfully used in the Surviving Sepsis Campaign [677], may also be effective. Suggested checklist items and patient management bundles are provided in Tables 4 and 5 of the original article.
Trauma care training should emphasize the critical role of coagulation in outcome. Enhancing clinician knowledge in this area should be integral to guideline implementation. Trauma centers should routinely evaluate performance using institutional quality management programs. Audits of best practice adherence, with feedback and practice changes as needed, should be part of local guideline implementation efforts. To assess care quality for bleeding trauma patients, adherence to the following quality standards is suggested:
- Time from injury to bleeding cessation intervention (surgery or embolization) in hypotensive patients unresponsive to initial resuscitation.
- Time from hospital arrival to full blood result availability (complete blood count, PT, fibrinogen, calcium, viscoelastic testing if available).
- Proportion of patients receiving TXA within 3 hours of injury.
- Time from hospital arrival to CT scan in bleeding patients without obvious hemorrhage source.
- Damage control surgical techniques used according to Recommendation 19.
- Thromboprophylaxis initiated according to Recommendation 37.
