INTRODUCTION — Massive transfusion is a treatment for massive hemorrhage. It can keep patients alive through volume and oxygen transport replacement while other methods of bleeding source control such as surgery, interventional radiology, endoscopy, uterine packing, or liver replacement are carried out, and then it can replace lost blood volume to facilitate recovery and healing.
Massive transfusion has been arbitrarily defined as the replacement by transfusion of ≥10 units of whole blood (WB) or red blood cells (RBCs) in 24 hours as an approximation of the replacement of at least one blood volume (see 'Definitions and jargon' below). It identifies groups of patients across multiple medical specialties with significant failure of vascular integrity and is also a situation in the transfusion service where the administrative constraints around issuing blood can delay patient care.
This topic discusses the epidemiology, practical aspects, and complications of massive transfusion, including an overview of the evolving Advanced Trauma Life Support approach to massive transfusion [1,2]. Clinical indications and common settings that require massive transfusion are discussed separately.
●Trauma – (See "Initial management of moderate to severe hemorrhage in the adult trauma patient" and "Ongoing assessment, monitoring, and resuscitation of the severely injured patient".)
●Cardiac surgery – (See "Achieving hemostasis after cardiac surgery with cardiopulmonary bypass".)
●Obstetric hemorrhage – (See "Postpartum hemorrhage: Medical and minimally invasive management".)
●Liver disease – (See "Anesthesia for the patient with liver disease".)
General information about transfusion thresholds and use of blood products is also presented separately. (See "Indications and hemoglobin thresholds for RBC transfusion in adults" and "Use of blood products in the critically ill".)
DEFINITIONS AND JARGON
●Massive transfusion – Massive transfusion was historically defined as transfusion of ≥10 units of whole blood (WB) or red blood cells (RBCs) in 24 hours.
This is an arbitrary definition. Patients who receive 9 units are not fundamentally different from those who receive 10, whereas those who receive 10 units over 30 minutes are very different from those who receive the same volume spaced over the course of a day.
Likewise, 10 units of RBCs in a child who weighs 30 pounds, which replaces their blood volume five times, is very different from 10 units of RBCs in an adult who weighs 300 pounds, which may only replace half of their original blood volume.
Patients who die of uncontrolled hemorrhage before they can receive 10 units of RBCs are missed by this definition, an argument for having massive hemorrhage protocols rather than massive transfusion protocols.
●Ultra-massive transfusion – Ultra-massive transfusion has been defined as using ≥20 units of RBCs in 24 or 48 hours [3]. It has been used to study the outcomes of patients who use very large amounts of blood resources in a short period of time.
Outcome is highly dependent on the clinical situation, survival being excellent after liver transplantation and very poor after gastrointestinal variceal bleeds. As a result, hyper-massive transfusion is not a useful measure or marker of clinical futility.
The definition has all the same problems as simple massive transfusion, being nonspecific and susceptible to changing patterns of care.
●Critical Administration Threshold for 1 hour (CAT-1) – CAT-1 is an alternative functional definition for acute catastrophic bleeding, in which the administration of ≥3 units of RBCs over one hour identifies patients with substantial immediate blood use [4]. Extension into subsequent hours of care (CAT-2, CAT-3, and so on) identifies patients whose hemorrhage control is not immediate and whose high-volume blood use is ongoing.
●Resuscitation Intensity (RI) Score – The RI score is similar to the CAT but uses ≥4 blood components (any components) in 30 minutes as a starting point and appends a number when more units are used (RI-6 or RI-12 and so on) [5]. The RI score is more sensitive in identifying patients needing rapidly issued or prepositioned blood products for treating shock and preventing rapid hemorrhagic death.
●Hemorrhage control resuscitation – Hemorrhage control resuscitation involves giving plasma and platelets early in resuscitation to avoid dilutional coagulopathy that occurred with the historic use of crystalloid fluids for volume resuscitation and RBCs in additive solution to maintain oxygen-carrying capacity [6].
First described by in a population of patients with ruptured abdominal aortic aneurisms, it typically involved a massive transfusion protocol (MTP) in which the blood bank issued coolers containing 4 units of RBCs, 4 units of plasma, and a pool of platelets, and with Cryoprecipitate and additional platelets guided by thromboelastography (TEG). The term is widely used in the European literature.
●Damage control resuscitation – Damage control resuscitation is the American military version of hemorrhage control resuscitation, with additional attention to hypothermia, acidosis, and fibrinolysis, as occurs in trauma-induced coagulopathy (TIC). Damage control resuscitation is the term typically used in the American trauma literature, either alone or in conjunction with damage control surgery.
Historically, damage control resuscitation used RBCs, plasma, and platelets in a 1:1:1 ratio when platelets were available, and fresh whole blood (WB) when they were not available in austere settings in the Southwest Asian military theater [7].
●1:1:1 – This describes the unit ratios of RBCs to plasma and platelets suggested in early papers for both hemorrhage control and damage control resuscitation.
The order of the components is not standardized. In the title of the Pragmatic Randomized Optimal Plasma and Platelets Ratio (PROPPR) trial paper, the ratios are given as 1:1:1 versus 1:1:2, listing the RBC fraction last. Generally, it would make more sense if the ratios had been given as 1:1:1 versus 1:0.5:0.5, listing the RBC fraction first and showing that relatively fewer units of plasma and platelets had been given in the second arm of the trial.
Another problem with 1:1:1 as a description of blood component ratios is that most platelet units in the United States are obtained by apheresis rather than prepared from WB donations. One apheresis unit is equivalent to five or six WB-derived units. The pooled units available in Canada and Europe have lower numbers of platelets and would have a different equivalency.
Most apheresis units available from the American Red Cross are pathogen-reduced and are associated with a 44 percent reduction in platelet recovery. Platelets in additive solution have had two-thirds of their plasma removed and replaced with a saline-based solution. It may be harder to achieve hemostatic concentrations of platelets and plasma coagulation factors using currently available components than in the past because of these changes. Benefits and limitations are discussed separately. (See "Pathogen inactivation of blood products", section on 'General principles of pathogen inactivation'.)
●Coagulopathy — Coagulopathy is a general term for reduced hemostasis caused by reduced platelet number or function, reduced coagulation factor concentrations or function, other interference with coagulation or platelet function, or accelerated clot breakdown.
Trauma-associated coagulopathy describes a constellation of findings in trauma patients. (See 'Trauma-associated coagulopathy' below.)
Simple measures of coagulation such as the platelet count and prothrombin time with international normalization ratio (PT/INR) are strong linear predictors of death in trauma across their normal range and the full range of injury severity. Coagulopathy is thus a relative term to be considered in the balance of burden of injury and coagulation capacity. (See 'Hemostatic abnormalities' below.)
Appropriate care requires knowing the blood products available at the local facility and how to put them together to achieve therapeutic goals.
EPIDEMIOLOGY — The need for massive transfusion depends on the patient population.
In a 2016 review of all 97,000 episodes of transfusion of ≥10 red blood cell (RBC) units in consecutive two-day periods over 25 years in two countries (Sweden, population 10 million, and Denmark, population 5 million), the most common clinical situations leading to massive transfusion were cardiac and vascular surgery accounting for almost 40 percent of all massive transfusions [3].
Other situations leading to massive transfusion, such as gastrointestinal hemorrhage, liver transplant, trauma, and obstetric catastrophes, were less frequent in the 2016 review [3]. While trauma and obstetric bleeding are considered classic causes of massive transfusion, they represented only 15 and 2 percent of all such episodes in the Swedish/Danish experience.
Indications were similar in a study involving 1300 episodes of ultra-massive transfusion of ≥20 units of RBCs in consecutive two-day periods in six hospitals in four countries [8].
Trauma is the most studied example from among indications for massive transfusion.
●In a 2021 review of experience in a major trauma center during the years 2011 to 2018 involving 47,471 patients, 18 percent received RBC transfusions, 2 percent received >10 units of RBCs in 24 hours, and 1 percent received >20 units in the first 24 hours [9].
●A 2012 multiple trauma center study reported that the fraction of all patients receiving RBC transfusions who received ≥10 units of RBCs has decreased by 40 percent, coinciding with the practice of earlier transfusion of plasma and platelet units [10].
Taken together, these data suggest that most injured patients never need massive transfusion and can be treated based on physical assessment and results of early laboratory tests. However, approximately 6 percent of the most severely injured individuals will require activation of a massive transfusion protocol, 2 percent will receive a massive transfusion, and 1 percent will receive ultra-massive transfusion. Appropriate rapid response has led to reduced blood use.
DATA FOR SPECIFIC PATIENT POPULATIONS
Trauma — Approximately 6 percent of trauma patients appear to be at risk of requiring massive transfusion based on vital signs, physical findings, or mechanism of injury. (See 'Epidemiology' above.)
Risk of massive hemorrhage is high in the following groups:
●Adults with systolic blood pressure (SBP) <70 mm Hg
●Adults with pulse >110 per minute and SBP <90 mm Hg
●Penetrating injury or major fractures
●Fluid in body cavities likely to be blood
Replacement with plasma, platelets, and red blood cells (RBCs) or whole blood (WB) should be started early. Extrapolation of observational studies carried out in the combat setting has driven significant interest in examining the impact of fixed transfusion ratios of plasma to platelets to RBCs on patient outcomes.
Results from several observational studies suggested that patients with severe trauma and coagulopathy requiring massive blood replacement were more frequently and successfully resuscitated when the ratio of transfused units of plasma to platelets to RBCs approached 1:1:1 [6,11-18]. In evaluating these studies it is important to remember the problem of survival bias, meaning that patients may have died because they did not receive plasma, or they may not have received plasma because they died [19]. This 1:1:1 resuscitation approach is referred to as "damage control resuscitation". (See 'Definitions and jargon' above and "Initial management of moderate to severe hemorrhage in the adult trauma patient".)
Observational studies were the impetus for a large randomized control trial in 680 trauma patients (the PROPPR [Pragmatic Randomized Optimal Platelet and Plasma Ratios] trial), which evaluated the outcomes of blood resuscitation protocols using a ratio of 1:1:1 (1 unit of RBCs for each unit of plasma and platelets) versus 1:1:2 (2 units of RBCs for each unit of plasma and platelets) [20]. This was the first large randomized trial in blood-product-based trauma resuscitation ever sponsored by the National Institutes of Health (NIH), and, in retrospect, it was significantly underpowered.
The PROPPR trial found no statistically significant difference in 24-hour and 30-day mortality with the 1:1:1 versus the 1:1:2 component ratios [20]. Planned secondary end-points showed higher rates of anatomic hemostasis and fewer deaths due to exsanguination at 24 hours in participants assigned to the 1:1:1 arm. However, as noted in a separate commentary, death by exsanguination alone was combined with death by exsanguination and other causes, which possibly overstated the impact of benefit of the 1:1:1 approach [21,22].
Although the benefit of a fixed 1:1:1 ratio resuscitation approach in massively bleeding trauma patients lacks the highest level of quality evidence, the approach has been widely adopted clinically and applied to trauma as well as other massive bleeding situations including obstetric, vascular, and liver transplant. Hemostatic resuscitation has been associated with reduced blood use and fewer deaths in these health care settings. However, the reason 1:1:1 therapy was adopted rapidly by the surgical community was that it avoided the major problem of crystalloid resuscitation, with massive tissue edema leading to delayed wound closure and prolonged ventilator requirements, and it directly addressed the problem of coagulopathy. Limiting crystalloid use in turn translates into fewer ventilator days and shorter hospital stays. It has become part of the National Academy of Medicine's "Zero Preventable Trauma Deaths" program [23]. (See 'Component ratio (1:1:1)' below.)
Additional aspects of transfusion and other interventions for trauma management are discussed in detail separately. (See "Initial management of moderate to severe hemorrhage in the adult trauma patient".)
Cardiac surgery — Cardiac surgery is the most common use of massive transfusion. However, even in patients with a substantial burden of surgery-related injury such as those undergoing repeat midline sternotomy and multivessel grafting or valve replacement, in the Red Cell Storage Duration Study (RECESS), only 145 of 1098 participants (13 percent) received ≥8 units of RBCs [24].
In a subset analysis involving 324 cardiac surgery patients who participated in RECESS, those receiving >5 units of RBCs had better outcomes (survival, preservation of organ function) when they received higher ratios of plasma and platelets to RBCs [24]. Mortality at day 7 was 1.7 percent in those who received at least 5 units of plasma per 5 units of RBCs, versus 7.2 percent in those who received lower ratios of plasma to RBCs (p = 0.0318).
Based on these improved outcomes with balanced resuscitation, we suggest using the 1:1:1 ratio to treat active bleeding in these individuals, in conjunction with laboratory testing. (See 'Laboratory monitoring' below.)
Other details of management are presented separately. (See "Achieving hemostasis after cardiac surgery with cardiopulmonary bypass".)
Obstetric hemorrhage — Pregnancy and the postpartum period are hypercoagulable states with compensatory increased fibrinolysis. When vascular disruption leads to massive bleeding, those with low fibrinogen are at increased risk for bleeding [25]. Guidelines from the College Nationale Gynecologie et Obstetrics Francais (CNGOF) suggest maintaining the fibrinogen ≥200 mg/dL [26].
Expert opinion from multiple groups reflects the data for trauma and suggest use of a 1:1:1 ratio for treating massive hemorrhage in obstetric patients.
Other details of management are presented separately. (See "Approach to the adult with vaginal bleeding in the emergency department", section on 'Initial management of unstable patient' and "Postpartum hemorrhage: Medical and minimally invasive management".)
Liver disease — Liver disease leads to reduced production of normal coagulation factors as well as dysfunctional vitamin K-dependent factors and dysfunctional fibrinogen. This is compounded by reduced hepatic clearance of activation fragments of coagulation factors, which can act as competitive inhibitors of coagulation enzymes [27]. Anticoagulant factors such as protein S and protein C are also reduced. As a result, laboratory values are difficult to interpret, and the effectiveness of conventional therapies may be reduced. (See "Hemostatic abnormalities in patients with liver disease".)
In patients with severe liver disease and active hemorrhage, attempts to normalize platelet number and plasma coagulation factor concentrations are probably helpful, with the caveats that increased blood volume will increase portal pressure, and increasing platelet counts in the face of splenomegaly is difficult. In liver transplant, balanced transfusion ratios may be helpful [28]. (See "Hemostatic abnormalities in patients with liver disease", section on 'Bleeding'.)
We suggest a 1:1:1 ratio of blood components for treating massive hemorrhage in individuals with liver disease [29].
Additional details of management are presented separately. (See "Liver transplantation: Anesthetic management", section on 'Blood and coagulation management' and "Hemostatic abnormalities in patients with liver disease", section on 'Bleeding'.)
APPROACH TO VOLUME AND BLOOD REPLACEMENT — Management of a patient requiring massive transfusion requires careful and ongoing consideration of complex physiologic relationships. The primary concerns are maintaining cardiac output, oxygen-carrying capacity, and hemostatic potential. There is no clear threshold for hematocrit, platelet count, or coagulation factor deficiency below which blood use is futile.
Crystalloid versus blood products — Correction of a deficit in blood volume with crystalloid volume expanders works well for most mildly and moderately ill or injured patients. However, large volume resuscitation with crystalloid alone in severe trauma with massive blood loss can lead to dilutional coagulopathy as well as severe tissue swelling, with stiff lungs and abdominal compartment syndrome.
For more severely injured patients, there has been a shift towards use of blood components for volume resuscitation (along with topical hemostatic agents), with avoidance of aggressive crystalloid resuscitation.
Attempts to treat dilutional coagulopathy once it is established run into the problem that, other than packed red blood cells (pRBCs), blood components themselves are already dilute or do not deliver high concentrations of blood cells and plasma proteins efficiently. Composition of various blood components is summarized in the table (table 1).
Component ratio (1:1:1) — For massive transfusion in most patient populations, a 1:1:1 ratio of plasma to platelets to pRBCs is generally used. Supportive evidence is discussed above. (See 'Trauma' above.)
●Plasma – 1 unit (as frozen plasma [FP], thawed plasma, or other plasma product) has a volume of 200 to 300 mL. (See "Clinical use of plasma components", section on 'Plasma products'.)
●Platelets – 1 unit of apheresis platelets contains 300 billion platelets, one-fourth of what normally circulates and one-sixth of what is in the body (one-third of platelets are in the spleen), but only half of transfused platelets typically circulate.
●pRBCs – 1 unit of pRBCs has a volume of approximately 325 mL and contains 160 to 220 mL of red blood cells (RBCs). (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion", section on 'Collection and storage procedures'.)
●Whole blood (WB) – WB units, available in some settings, can deliver all the components in normal ratios without the dilution and saline load from the RBC additive solution [30,31].
Blood warmer — A blood warmer should be used to prevent hypothermia, which can exacerbate other complications. (See 'Hypothermia' below.)
Rationale for target hemoglobin — The American Society of Anesthesiologists recommends avoiding hemoglobin <6 g/dL in healthy individuals. For older and sicker individuals with cardiovascular disease, a hemoglobin of ≥8 g/dL appears reasonable [32]. Supporting data are presented separately. (See "Indications and hemoglobin thresholds for RBC transfusion in adults", section on 'Rationale for transfusion'.)
There is also a biologic rationale for a target hemoglobin. At rest, oxygen delivery is normally four times oxygen consumption, indicating the presence of an enormous reserve. If intravascular volume is maintained during bleeding and cardiovascular status is not impaired, oxygen delivery will theoretically be adequate until the hematocrit (packed cell volume) falls to 10 to 12 percent. This is because adequate cardiac output plus increased oxygen extraction can compensate for the decrease in arterial oxygen content.
However, increasing cardiac work to increase output requires more oxygen, so the "critical point" where oxygen consumption becomes delivery-dependent is generally higher. Transfusion of pRBCs delivered at 1:1:1 have a hematocrit in the total delivered volume of 29 percent, so there is "room" for delivered drugs, and non-RBC containing products such as platelets and cryoprecipitate.
LABORATORY MONITORING — Transfusions (with a 1:1:1 ratio) are used initially; laboratory monitoring should be performed over the course of resuscitation to further guide therapy.
Hemoglobin and hemostatic testing — The following tests are monitored to assess hemoglobin and hemostatic capacity, preferably after each 5 units of packed RBCs (pRBCs) or as the clinical situation allows:
●Complete blood count (CBC) with platelet count
●Prothrombin time (PT)
●Activated partial thromboplastin time (aPTT)
●Fibrinogen concentration
If viscoelastic testing is immediately available and clinicians are familiar with its use, it can be used instead of the PT, aPTT, and fibrinogen, but viscoelastic tests are less sensitive, less specific, slower, and more expensive [33-35].
The two main tests are (table 2):
●Thromboelastography (TEG)
●Rotational thromboelastography (ROTEM)
These tests can be useful to confirm the dilutional effects of transfusion and their prevention by appropriate component ratios, as well as development of other complications such as disseminated intravascular coagulation (DIC). (See 'Hemostatic abnormalities' below.)
Although the CBC, PT, aPTT, and fibrinogen are widely used, it is well recognized that laboratory-based assays may not provide real-time assessment of a patient's hemostatic status due to the rapid evolution of the clinical status of a massively bleeding patient.
Cognizant of long turnaround times of laboratory-based assays, clinical laboratories have adopted protocols whereby the turnaround time is shortened for key laboratory parameters (hemoglobin, platelet count, PT, fibrinogen) through adjustment of preanalytic centrifugation times and shortening of the von Claus fibrinogen monitoring time to 10 minutes. Building such an emergency hemorrhage panel with a turnaround time of 15 to 20 minutes is a reasonable goal.
Alternatively, some institutions use point-of-care platforms for coagulation testing or viscoelastic testing in lieu of laboratory-based tests to expedite hemostatic assessment and improve blood component usage.
Details of these tests are presented separately. (See "Clinical use of coagulation tests", section on 'Clotting times' and "Clinical use of coagulation tests", section on 'Fibrinogen' and "Clinical use of coagulation tests", section on 'Point-of-care testing'.)
Monitoring pH, blood gases, electrolytes, and metabolic parameters — Early and frequent measures of pH, blood gases, electrolytes, and metabolites such as glucose and lactate provide a wealth of information early in critical care and throughout massive transfusion. Testing can be done on arterial or venous blood. We perform these measurements on a calibrated blood gas machine every 20 to 30 minutes while patients are being actively resuscitated.
COMPLICATIONS — High volume and rapid transfusions are associated with a number of hemostatic and metabolic complications [36]. These can be minimized by selecting appropriate types of blood components and considering volume status, tissue oxygenation, bleeding, and metabolic parameters (calcium, potassium, and acid-base).
Hemostatic abnormalities — Coagulopathy can be common in massive transfusion. Contributing factors include clotting factor consumption and activation due to tissue trauma (massive head injury or muscle damage) or reduced clotting factor activity due to dilution, prolonged shock, hypoxia-induced acidosis, hypothermia, or failure to clear activation peptides that act as competitive inhibitors [37]. (See 'Dilutional coagulopathy' below and 'Trauma-associated coagulopathy' below and 'Impaired platelet and coagulation factor function' below and 'Hypothermia' below.)
Dilutional coagulopathy
Decreased coagulation factors and platelets — Coagulation abnormalities may be induced by packed red blood cell (pRBC) transfusions or crystalloid infusions, which dilute coagulation proteins and platelets [38-40]. pRBCs are essentially devoid of plasma and platelets.
Gradual dilution of clotting proteins leads to prolongation of the prothrombin time (PT) and activated partial thromboplastin time (aPTT). In an adult, there will be an approximate 10 percent decrease in the concentration of clotting proteins for each 500 mL of blood loss that is replaced with plasma-poor pRBCs in additive solution. Additional bleeding based solely on dilution can occur when the level of individual coagulation proteins falls to <25 percent of normal. This usually requires 6 to 10 units of pRBCs in an adult.
A similar dilutional effect on the platelet concentration can be seen with massive transfusion [41]. In an adult, each 10 to 12 units of transfused pRBCs are associated with a 50 percent fall in the platelet count, and significant thrombocytopenia can be seen after 10 to 20 units of blood, with platelet counts <50,000/microL.
Plasma, fibrinogen, and platelet transfusions for dilutional coagulopathy — Monitoring is discussed above. (see 'Hemoglobin and hemostatic testing' above).
The following is appropriate to correct dilutional coagulopathy:
●Plasma – Plasma contains all the clotting factors. If the PT and/or aPTT are >1.5 times control due to dilutional coagulopathy, 2 to 8 units of plasma should be given. Each unit of plasma might be expected to increase the clotting protein levels by 6 percent in an adult, but because of losses in product preparation, storage, and of transfused factors to the interstitial space, typical increments are of the order of 2.5 percent [42]. (See "Clinical use of plasma components", section on 'Plasma products'.)
●Fibrinogen – Cryoprecipitate or virally inactivated fibrinogen concentrate may be used when fibrinogen levels are critically low (<100 mg/dL) [1]. (See "Disorders of fibrinogen", section on 'Treatment/prevention of bleeding'.)
●Platelets – Platelet transfusions (six units of whole blood [WB]-derived platelets or one apheresis unit) should be given if the platelet count is <50,000/microL. Each unit should increase the platelet count by 5000/microL (30,000/microL for a full six-unit adult dose). (See "Platelet transfusion: Indications, ordering, and associated risks", section on 'Massive blood loss'.)
Trauma-associated coagulopathy — Trauma-associated coagulopathy can be diagnosed when there is microvascular oozing, prolongation of the PT and aPTT more than expected by dilution, significant thrombocytopenia, low fibrinogen levels, and increased D-dimer levels. This is sometimes referred to as "traumatic disseminated intravascular coagulation" (traumatic DIC), although this is probably a misnomer, as there are several differences from DIC. (See "Evaluation and management of disseminated intravascular coagulation (DIC) in adults", section on 'Acute DIC'.)
Trauma-associated coagulopathy:
●Is not truly disseminated (it exists at multiple sites of endothelial disruption)
●Is not truly intravascular (it is driven by extravascular tissue factor and subendothelial matrix collagen)
●Is not truly coagulation (it is bleeding)
●Is effectively treated with antifibrinolytic agents, which are contraindicated in true DIC
No single laboratory test can definitively diagnose or exclude trauma-associated coagulopathy, and laboratory correlation with the patient's clinical picture is a priority.
Impaired platelet and coagulation factor function — Acidosis and hypothermia can both interfere with the normal clotting and hemostasis:
●Acidosis - Acidosis (excess protons) interferes with the assembly of coagulation factor complexes involving calcium and negatively charged phospholipids (clustered as "rafts" on the surfaces of platelets, cells expressing tissue factor, and damaged endothelial cells). As a result, the activity of the factor Xa/Va prothrombinase complex at a pH of 7.2, 7.0, and 6.8 is reduced by 50, 70, and 80 percent, respectively [43].
The resulting delayed production and reduced concentrations of generated thrombin lead to delayed fibrin production, altered fibrin structure, and increased susceptibility to fibrinolysis [44]. The contribution of acidosis with coagulopathy to increased trauma mortality has been widely recognized. (See "Etiology and diagnosis of coagulopathy in trauma patients", section on 'Acidosis'.)
●Hypothermia – Hypothermia reduces the enzymatic activity of plasma coagulation proteins; it has a greater effect on hemostasis by preventing the activation of platelets via traction on the glycoprotein Ib/IX/V complex by von Willebrand factor [45]. The onset of this effect is seen at core temperatures of ≤34°C.
In tests of shear-dependent platelet activation, this pathway stops functioning at 30°C in 50 percent of individuals and is markedly diminished in most of the rest. This profound effect on platelet-mediated primary hemostasis means that massive bleeding in conjunction with a core temperature of <30°C was rarely survived in the past [46]. Prevention and treatment of hypothermia are critical. (See 'Hypothermia' below.)
Metabolic abnormalities
Hypocalcemia from citrate toxicity — Hypocalcemia is a common finding with unclear clinical significance.
Half of hospitalized patients have low ionized calcium when compared to healthy volunteers, and 70 to 100 percent of trauma patients have low calcium on admission before blood is administered [47,48]. Treatment of mild or moderate hypocalcemia does not appear to change outcome.
Blood is anticoagulated with sodium citrate and citric acid [49]. Each 450 mL unit of blood (or 500 mL unit in Europe) contains 9 mmol of citrate. As a result, massive transfusion leads to infusion of a large amount of citrate. Infused citrate can lead to a clinically significant fall in the plasma free calcium (ionized calcium) concentration. This change can lead to paresthesias and/or cardiac arrhythmias in some patients [50]. (See "Relation between total and ionized serum calcium concentrations" and "Clinical manifestations of hypocalcemia", section on 'Acute manifestations'.)
●Normal calcium ranges and expected changes – Ionized calcium in healthy adults ranges from 1.18 to 1.38 mmol/L. Half of hospitalized patients in whom ionized calcium is measured have ionized calcium levels below this normal concentration, as do 80 to 100 percent of trauma patients at the time of admission, before any blood is administered [47,48].
Administering whole blood (WB) at 100 mL/min lowers the ionized calcium to approximately 0.7 mmol/L, with a half-time of <2 minutes, and the ionized calcium returns to normal when the transfusion is stopped, with the same kinetics [51]. Some people become symptomatic (or develop clinical findings) when blood is transfused or apheresis procedures are performed at this volume and rate, and some do not.
Individuals undergoing apheresis with a plasma exchange rate at 50 to 100 mL/min are frequently treated with supplemental calcium at 4 mmol/hour, with symptomatic control. (See "Therapeutic apheresis (plasma exchange or cytapheresis): Complications", section on 'Hypocalcemia'.)
●Prevention – The risk of hypocalcemia can be reduced by paying attention to the rate of transfusion and following physiologic markers such as the blood pressure and corrected QT interval. Measuring the ionized calcium concentration with the ion-specific electrode on a blood gas machine is common. Monitoring is discussed above. (See 'Monitoring pH, blood gases, electrolytes, and metabolic parameters' above.)
Citrate is metabolized by the liver. Individuals with underlying liver disease or ischemia-induced hepatic dysfunction are at higher risk of citrate-induced hypocalcemia. In patients with liver disease or decreased hepatic perfusion, ionized calcium should be monitored.
●Treatment – Most patients do not require calcium administration. In a retrospective review of 348 consecutive trauma patients with ionized calcium as low as 0.5 mmol/L, there were no differences in outcomes for those who were treated with calcium and those who were not treated [48].
However, calcium should be administered to treat symptomatic hypocalcemia, with ongoing monitoring and attention paid to not administering too much calcium and causing hypercalcemia.
Many individuals receiving massive transfusion are under anesthesia and are not able to report symptoms, and strong evidence supporting treatment is lacking. Thus, the blood pressure and corrected QT interval may be the only findings that alert the clinician to the presence of severe, symptomatic hypocalcemia that should be treated.
Calcium can be administered as calcium chloride or calcium gluconate. For individuals with abnormal liver function, calcium chloride may be preferable to calcium gluconate since release of ionized calcium from calcium chloride does not require normal liver function [1]. Ionized calcium will be released from calcium gluconate more slowly if liver function is abnormal.
Intravenous dosing is suggested as follows:
•Calcium chloride – Give 2 to 5 mL (up to 10 mL) of a 10 percent calcium chloride solution (1.6 to 3.9 mmols, up to 7.8 mmol) per unit of blood (RBCS and plasma).
•Calcium gluconate – Give 10 to 20 mL of a 10 percent calcium gluconate solution (2.5 to 5.0 mmols) per unit of blood.
The patient should be assessed for clinical effect and the serum calcium should be monitored.
●Reasons not to treat – By extrapolation from animal studies, it is possible to calculate the maximum transfusion rate that would permit a healthy liver to metabolize excess citrate, thereby avoiding hypocalcemia. The maximum citrate infusion rate should be 0.02 mmol/kg per minute (since this represents the maximum rate of citrate metabolism) and the citrate concentration in WB is 20 mmol/L (0.020 mmol/mL). Thus:
Maximum citrate infusion rate (mmol/kg per min) = (mmol citrate per mL of blood x mL of blood infused per min) ÷ wt (kg)
mL of blood infused per min = (0.02 ÷ 0.015) x wt (kg) = 1.33 x wt (kg)
For a 50 kg recipient with normal liver function and adequate liver perfusion, the maximum rate of blood transfusion to avoid citrate toxicity is 66.5 mL/min, equivalent to approximately 9 units of WB per hour (450 mL per unit). Significant hypocalcemia should not develop with slower rates except under extreme circumstances. However, the risk is substantially greater in a patient with preexisting liver disease or ischemia-induced hepatic dysfunction. In such patients, the plasma ionized calcium concentration should be monitored. (See 'Monitoring pH, blood gases, electrolytes, and metabolic parameters' above.)
Hyperkalemia — (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion", section on 'Hyperkalemia'.)
During storage, potassium leakage causes the supernatant of RBC units to increase by 1 mEq/day, increasing from approximately 3 mEq/L at the time of donation to 45 mEq/L during 42 days of storage. Irradiation can increase this rate to 1.5 mEq/day.
Nevertheless, because the volume of suspending additive solution in a unit of pRBCs is small (150 mL), the actual amount of free potassium infused with a unit of RBC is only approximately 7 mEq/unit. If blood high in extracellular potassium is infused directly into the heart, it can cause arrythmias. This is usually only a problem with high volume transfusion or blood exchange with heart lung machines, extracorporeal membrane oxygenation (ECMO) devices, and apheresis machines primed with older, cold-stored blood.
In these patients, the following steps can be used to minimize the risk of hyperkalemia:
●Use younger RBCs – Select only RBCs collected <10 days prior to transfusion.
●Use washed RBCs – Any unit of RBCs can be washed immediately before infusion to remove extracellular potassium.
In other patients receiving massive transfusion, hyperkalemia is generally not a problem, since the potassium in the supernatant is only approximately 7 mEq, and this is rapidly pumped back into the RBCs in the transfusion as the cells warm.
Hypothermia — Hypothermia can exacerbate bleeding, cause arrythmias, and lead to other complications [52]. (See 'Impaired platelet and coagulation factor function' above.)
A high-capacity commercial blood warmer should be used to warm blood components toward body temperature when >2 units are transfused. (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion", section on 'Pre-transfusion considerations'.)
Rapid transfusion of multiple units of chilled blood may reduce the core temperature abruptly. The human body has a specific heat of approximately 0.86.
●Six units (or two liters) of blood at 4°C will reduce the body temperature of a 70 kg individual by approximately 1°C.
●Heat loss can be additive with evaporative heat loss associated with an open abdomen or other body cavity which, by itself, can lead to a 1°C decrease in core temperature in 40 minutes.
●Ten units of cold blood products and an hour of surgery can lead to a 3°C drop in core temperature and hypothermic coagulopathy. (See 'Impaired platelet and coagulation factor function' above.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Transfusion and patient blood management".)
SUMMARY AND RECOMMENDATIONS
●Definition – Massive transfusion has been defined as transfusion of ≥10 units of whole blood (WB) or packed red blood cells (pRBCs) in 24 hours, ≥3 units of pRBCs in one hour, or ≥4 blood components in 30 minutes, recognizing that blood loss is a continuum and these are arbitrary cutoffs. It identifies patients who require ongoing considerations of complex physiological relationships related to cardiac output, oxygen-carrying capacity, and hemostasis. (See 'Definitions and jargon' above.)
●Indications – Common indications for massive transfusion include trauma, cardiac surgery, obstetric bleeding, and liver disease. Cardiac surgery is the most common; trauma is the most extensively studied. Approximately 3 percent of the most severely injured individuals will require massive transfusion. Data from randomized trials in trauma patients remain inconclusive on the best approach, but approaches tested in trauma patients have been widely adopted. (See 'Epidemiology' above and 'Data for specific patient populations' above.)
●Administration – Crystalloid volume expanders generally work well to correct a volume deficit for most mildly and moderately ill or injured patients. For more severely injured patients, practice has shifted to use of blood component resuscitation. Volumes are summarized in the table (table 1). (See 'Approach to volume and blood replacement' above.)
•For patients who require massive (or ultra-massive) transfusion (>5 units of RBCs in a day or >3 units in 2 hours), we suggest blood component resuscitation rather than crystalloid (Grade 2C). (See 'Crystalloid versus blood products' above.)
•For individuals treated with blood component resuscitation, we suggest a ratio of 1:1:1 (plasma to platelets to pRBCs) rather than other ratios (Grade 2C). (See 'Component ratio (1:1:1)' above.)
•Hemoglobin <6 g/dL should be avoided (<7 or 8 g/dL in selected individuals). (See 'Rationale for target hemoglobin' above.)
●Monitoring – Attention must be paid to hemoglobin, platelet count, hemostasis, and metabolic status. This includes complete blood count (CBC) with platelet count after each 5 units, along with coagulation testing. Standard tests of coagulation include the prothrombin time (PT), activated partial thromboplastin time (aPTT), and fibrinogen concentration; if available and clinicians are familiar with use, a viscoelastic test such as thromboelastography (TEG) (table 2) can be used instead. Ionized calcium and other metabolic parameters should be monitored and treated systematically. (See 'Laboratory monitoring' above.)
●Complications
•Hemostatic – Plasma (2 to 8 units) is transfused if the PT or aPTT exceed 1.5 times the control value due to dilutional coagulopathy. Platelets (1 apheresis unit or 6 units of WB-derived platelets) are transfused if the platelet count decreases to <50,000/microL. Cryoprecipitate or fibrinogen concentrate can be used for fibrinogen <100 mg/dL (<200 for obstetric bleeding). (See 'Hemostatic abnormalities' above.)
•Metabolic – Hypocalcemia can be caused by citrate toxicity; symptomatic hypocalcemia is treated with intravenous calcium chloride or calcium gluconate (2 to 5 mL of a 10 percent calcium chloride solution per unit of WB or pRBCs plus plasma; maximum 10 mL). Acid-base abnormalities and hyperkalemia should be treated if present.
•Hypothermia – Hypothermia can exacerbate bleeding, cause arrythmias, and lead to other complications. A high-capacity commercial blood warmer should be used to warm blood components toward body temperature when >2 units are transfused. (See 'Hypothermia' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff gratefully acknowledges the extensive contributions of Arthur J Silvergleid, MD, to earlier versions of this and many other topic reviews.
Do you want to add Medilib to your home screen?