Chapter 29. Transfusion in Critical Care

Fresh, warm whole blood most effectively restores red cell mass, plasma volume, clotting factors, and platelets. However, given shortages of blood products, the use of whole blood transfusions is not realistic. The use of component product transfusion is the mainstay of blood banking and transfusion practice, effectively utilizing a scare resource while matching the components transfused to the specific needs of the patient. Whole blood is usually separated into packed red blood cells (PRBC), fresh frozen plasma (FFP), and platelet concentrate soon after donation. The plasma can be further separated into cryoprecipitate and cryopoor plasma, or undergo further fractionation to individual plasma proteins.

Indications for blood component therapy can be divided into two main categories: (1) enhancement of oxygen-carrying capacity by increasing red blood cell (RBC) mass and (2) replacement of coagulation components due to loss, dysfunction, or consumption.

Anemia is one of the most common abnormal laboratory findings among critically ill patients. The effect of anemia on outcome and the determination of transfusion triggers has been the subject of much debate in recent literature.

Historically, the decision to transfuse has been guided by the hemoglobin (Hb) concentration, usually 10 mg/dL. However, given the risks associated with PRBC transfusion and recent literature supporting better or similar outcomes with lower transfusion triggers, the optimal Hb level at which to transfuse patients remains unclear.

Benefits of RBC Transfusion

The main function of RBCs is to transport oxygen from the lungs to the peripheral tissues. Oxygen delivery (DO2) is calculated by multiplying the cardiac output (CO) times the arterial oxygen content (CaO2):

where DO2 is in milliliters per minute, CO in deciliters per minute, and CaO2 in milliliters per deciliter.

And CaO2 is calculated by the following equation:

where SaO2 is the arterial oxygen saturation (%), 1.34 the oxygen-carrying capacity of Hb (mL/g), [Hb] the Hb concentration (g/dL), 0.0031 the solubility of oxygen in plasma at 37°C, and Pao2 is measured in millimeters of mercury.

Under normal conditions, DO2 exceeds oxygen consumption (VO2) by three to five times. However, in situations where the VO2 of the peripheral tissues is greatly increased, or DO2 is decreased by anemia or decreased CO, VO2 can exceed DO2 and result in tissue hypoxia. Increasing the [Hb] is one of the ways to increase the blood's oxygen-carrying capacity, and therefore increase DO2. Additionally, transfusion can increase blood volume for patients following acute blood loss or hemorrhage and alleviate symptoms of anemia such as dyspnea, weakness, and fatigue.

Drawbacks of RBC Transfusion

Despite the theoretical benefits of transfusion described above, it is associated with multiple risks. There is risk of human error resulting in a transfusion reaction, most commonly an acute hemolytic reaction, from receiving incorrectly matched blood. Febrile reactions, namely, nonhemolytic/noninfectious reactions secondary to antileukocyte antibodies, are also possible in up to 7% of blood product recipients. There is risk of allergic reaction ranging from urticaria to frank anaphylaxis that is usually a result of the passive transfer of sensitizing antibodies. Additionally, transmission of communicable diseases such as human immunodeficiency virus (HIV) and viral hepatitis is possible, although this risk, with modern blood banking techniques, is now exceedingly remote.1 Finally, metabolic derangements can occur with transfusion, such as hypocalcemia and hyperkalemia.

More commonly, especially in the critically ill population, transfusion of PRBCs is associated with increased risk of infection, including wound infections, sepsis and pneumonia,2,3 increased incidence of multiple organ failure (MOF),4 and increased risk of acute lung injury (ALI) and acute respiratory distress syndrome (ARDS).5 In addition, transfusions are associated with longer intensive care unit (ICU) and hospital length of stay (LOS), more complications, and increased mortality.6 These effects are dose-dependent, meaning the more units of blood that are transfused, the higher the risk of complications.

The reason for the increased morbidity and mortality in patients receiving PRBCs is not completely clear, although recent interest has focused on immunomodulating effects of transfused RBCs and RBC storage lesions (age of transfused RBCs) as possible mechanisms. It has been suggested that leukodepleted blood may have less immunomodulating properties and, hence, reduce the complications associated with the transfusion of nonleukodepleted blood,7–9 but there is still considerable debate about the benefit of leukoreduction and which patients will benefit most from receiving leukoreduced blood.10 Similarly, age of transfused RBCs has also been suggested as a possible explanation for the adverse effects associated with RBC transfusion. Well-documented changes occur to the RBC product during ex vivo storage, including a reduction in RBC deformability, altered RBC adhesiveness and aggregability, and a reduction in 2,3-diphosphoglycerate and adenosine triphosphate (ATP). These changes reduce posttransfusion viability of RBCs and limit DO2.11 The clinical effect of these changes is uncertain; however, some studies have suggested that transfusion of “older” RBCs may be associated with adverse effects.12–14 But a 2009 review of 27 studies of postsurgical, ICU, and trauma patients could not establish definitive relationship between the age of transfused RBCs and outcome in adult patients, except possibly for trauma patients receiving massive transfusion.15

Transfusion Threshold

There have been multiple retrospective and observational studies that demonstrate that the use of blood transfusions for the treatment of anemia in hemodynamically stable critically ill patients is not associated with improved outcome. The CRIT Study, performed in the United States and published in 2004, documented that 44% of all patients received blood transfusions and the number of units transfused was independently associated with worse outcomes.16 Similarly, a European study by Vincent et al showed an ICU transfusion rate of 37%. In this study, the group of patients who received a transfusion had a higher mortality rate compared with the nontransfused group despite similar degrees of organ dysfunction.17

The Transfusion Requirements in Critical Care (TRICC) trial conducted in Canada is the only prospective, adequately powered study that randomized patients to either a restrictive transfusion strategy (patients transfused if Hb dropped below 7 mg/dL and maintained between 7 and 9 mg/dL) or a liberal strategy (patients transfused when Hb fell below 10 mg/dL and maintained at 10–12 mg/dL). The overall hospital mortality was significantly lower in the restrictive transfusion group (22.2% vs. 28.1%, P = .05), and although 30-day mortality was similar in the two groups (18.7% vs. 23.3%, P = .11), mortality rates were significantly lower in patients randomized to the restrictive transfusion group who were <55 years of age and less acutely ill. The authors concluded that a restrictive strategy of red cell transfusion is at least as effective as, and possibly superior to, a liberal transfusion threshold for hemodynamically stable critically ill adults.18 Despite these recommendations, multiple studies demonstrate that transfusion practices have not changed significantly.

One patient population that may be an exception to the restrictive transfusion threshold is elderly patients with evidence of myocardial ischemia. A group of 79,000 patients >65 years of age admitted to the hospital with myocardial infarction was retrospectively reviewed. Those patients with lower hematocrit (HCT) levels on admission had higher 30-day mortality. In addition, blood transfusion was associated with a lower 30-day mortality among those patients with an HCT <24%. The patients who had an HCT above 30% did not have improved mortality from transfusion.19 It is important to remember that this study was performed in patients with evidence of myocardial ischemia. A history of coronary artery disease (CAD) or patients at risk for ischemia do not fall into this category.20

Sepsis

The Surviving Sepsis Campaign recommends that during the first 6 hours of resuscitation of severe sepsis or septic shock, if central venous oxygen saturation of 70% is not achieved with fluid resuscitation to a central venous pressure (CVP) of 8–12 mm Hg, then transfuse PRBCs to an HCT >30% and/or initiate dobutamine infusion.21 This recommendation is based on the study of Rivers et al regarding early goal-directed therapy for sepsis.22 Once initial resuscitation of tissue hypoxia is achieved, and in the absence of myocardial ischemia, the restrictive transfusion threshold should then be targeted.

Trauma

The trauma patient with hemorrhagic shock should be transfused regardless of Hb levels. However, in the absence of ongoing blood loss or hemorrhagic shock, there is no benefit of a “liberal” transfusion threshold in hemodynamically stable trauma patients.23 In a prospective study of over 15,000 trauma patients, blood transfusion was shown to be an independent predictor of mortality, ICU admission, ICU LOS, and hospital LOS. Those patients who received blood transfusion in the first 24 hours were more than three times as likely to die.24

Traumatic Brain Injury

One area that remains unclear is transfusion criteria for those patients with traumatic brain injury (TBI). One retrospective review showed that more days with an HCT <30% were associated with improved neurologic outcome, but the lowest measured HCT was associated with lower Glasgow Coma Scale (GCS) score on hospital discharge. The authors concluded that patients with TBI should not have different transfusion thresholds than other critical care patients, but prospective studies need to be done to support this recommendation.25

Recommendations

A number of guidelines regarding the indications for RBC transfusion have been published between 1997 and 2007. Most recently, a joint task force of the Eastern Association for the Surgery of Trauma (EAST) and the American College of Critical Care Medicine (ACCM) of the Society of Critical Care Medicine (SCCM) performed a comprehensive literature review of the topic and graded the evidence using scientific assessment methods. Table 29-1 includes an abridged summary of their 2009 evidence-based recommendations regarding the use of RBC transfusion in adult trauma and critical care.26

Anemia in critical illness is extremely common, and up to 40% of ICU patients will be transfused during their hospital stay. Physicians must weigh the risks and benefits of transfusion. PRBC transfusions are associated with increased incidence of infection, MOF, ALI, and ARDS across heterogeneous patient groups. Based on existing literature, no one transfusion trigger should be used in all patients. However, the evidence is sufficient to state that transfusions are rarely beneficial when Hb level exceeds 10 g/dL (HCT >30%) in the absence of acute blood loss, and using a restrictive strategy for transfusion (transfusing PRBC when Hb levels fall below 7 mg/dL) is just as effective, and likely superior to, a liberal strategy in hemodynamically stable critically ill patients.

FFP is plasma separated from the RBCs and platelets of whole blood and placed at −18°C or below within 8 hours after collection. By definition, “1 U” of FFP has the equivalent plasma coagulation factors as 1 U of whole blood and one bag contains approximately 200–250 mL. Once thawed, FFP must be used within 24 hours or the amount of factors V and III begins to decline. FFP is not a concentrate and must be ABO compatible.

Indications

Transfusion of FFP is indicated in the presence of active bleeding or prior to major invasive procedures with known or suspected coagulation abnormalities due to inadequate production, malfunction, loss, or consumption of multiple clotting factors.27–30 Liver failure, warfarin overdose, vitamin K deficiency, and dilutional coagulopathy are some indications for plasma transfusion. Patients with a deficiency of a single clotting factor are more appropriately treated with factor concentrates or cryoprecipitate.

Coagulopathy is supported by prothrombin time (PT) or international normalized ratio (INR) greater than 1.5 times normal, or activated partial thromboplastin time (aPTT) greater than 1.5 times the top of the normal range.31 Even in the setting of excessive warfarin effects, transfusion of plasma products should not be used to reverse elevated INR in the absence of bleeding, unless urgent invasive or surgical procedures are required.32

Mild to Moderate Coagulopathy

The ability of FFP to reverse mild to moderate coagulopathy (INR 1.1–2) has been shown to be poor.33,34 Regardless of the number of units of FFP transfused, there is a low likelihood that the INR will be corrected to normal levels.35 Specifically, current evidence does not support the use of prophylactic plasma transfusion for minimally invasive procedures in the setting of mildly abnormal coagulation tests such as paracentesis, thoracentesis,28 or central venous catheter insertion.36

Approximately 25% of clotting activity is required for hemostasis. Given that the plasma volume of humans is usually 40 mL/kg, the amount required is approximately 10–15 mL/kg, or 2–3 U of FFP, in the absence of ongoing losses or consumption. This is a general guideline, and clinicians should follow clinical course and coagulation parameters to guide transfusion, remembering that mild to moderate coagulopathy may not be corrected with FFP.

Massive Transfusion

Prophylactic transfusion of FFP is indicated in patients receiving massive transfusion (defined as greater than 10 U of PRBC in a 24-hour period). The exact ratio of FFP to PRBC transfusion has been a debate for many years. Historically, the FFP:PRBC ratio ranged from 1:4 to 1:10 and the initiation of almost any massive transfusion protocol using multiple different ratios had been shown to improve mortality.37 Recent literature in both military and civilian trauma patients supports using higher ratios of FFP. The optimal ratio appears to be between 1:1 and 1:3 FFP to PRBC38–40 and is a source of continued study.

Cryoprecipitate is obtained from the precipitate of frozen plasma once it is thawed. It has high concentrates of factor VIII, fibrinogen, factor XIII, and von Willebrand factor but the volume of cryoprecipitate is smaller, approximately 10 mL, and multiple units are often combined for transfusion. Despite the small volume of cryoprecipitate, it carries the same infectious risk as 1 U of FFP. Indications for transfusion of cryoprecipitate include fibrinogen deficiency with levels less than 100 mg/dL, mostly encountered during massive bleeding or consumptive coagulopathy; von Willebrand disease; and hemophilia A when factor VIII concentrates are not available.31

Platelets are required for primary hemostasis and circulate normally at a count of 150–400 × 109/L. Each concentrate of platelets contains approximately 5.5 × 1010 platelets and is derived from 1 U of whole blood or from plateletpheresis donations. Once collected, platelets can be stored for up to 5 days, and they are pooled with concentrates from multiple donors before transfusion. ABO compatibility is not required, but is preferred because small amounts of donor leukocytes and plasma are transfused along with the platelets. Each unit of platelets is expected to increase the platelet count by 5–10 × 109/L in the absence of consumption or ongoing loss, and the usual dose is 1 U per 10 kg of body weight.

There is no single target platelet count under which transfusion is recommended for all patients. When platelet counts fall below 5 × 109/L, there is a possibility of spontaneous hemorrhage and a high risk of hemorrhage with trauma or an invasive procedure.41 Given these risks, platelets should be administered regardless of apparent bleeding, when platelets drop below this level.42 At counts greater than 50 × 109/L, bleeding due to platelet deficiency is unlikely and prophylactic transfusion is normally not indicated. For patients with active bleeding and those undergoing invasive or surgical procedures, the current recommendation is that platelet counts should be maintained above 50 × 109/L.31 Some recommend targets of 100 × 109/L in the setting of intracranial hemorrhage or multisystem trauma.43

Platelet counts between 5 × 109 and 50 × 109/L have variable risks of hemorrhage due to thrombocytopenia, and the issue of prophylactic platelet transfusion is controversial between these levels. Clinical observation and evaluation of the patient's other risk factors for bleeding must guide transfusion practices. Prior recommendations were to transfuse platelets whenever the platelet count was less than 20 × 109/L, but recent literature recommends lowering that trigger to 10 × 109/L.44,45

Patients suffering from a destructive cause for thrombocytopenia rarely benefit from platelet transfusion because the transfused platelets are rapidly destroyed. Patients with such conditions as idiopathic thrombocytopenia purpura (ITP), hypersplenism, disseminated intravascular coagulation (DIC), sepsis, platelet antibodies, or those after cardiac surgery with extracorporeal bypass fall under this category. In the presence of a life-threatening hemorrhage or surgery, transfusion may be beneficial for its short-term effect. Platelet transfusion is contraindicated in such conditions as thrombotic thrombocytopenic purpura (TTP)46 and hemolytic uremic syndrome (HUS) because of worse outcomes, and is therefore reserved for life-threatening hemorrhage with these conditions. Heparin-induced thrombocytopenia (HIT) was also thought to be a contraindication for platelet transfusion; however, recent guidelines conclude that platelet transfusion can be considered in patients with HIT and overt bleeding, or those thought to be at high risk of bleeding.47

The use of recombinant erythropoietin (EPO) has been shown to reduce the need for RBC transfusions in patients with chronic renal failure, as well as those with anemia of chronic disease, such as cancer and acquired immune deficiency syndrome (AIDS).48 Despite the decrease in production of endogenous EPO that accompanies critical illness,49 the use of recombinant EPO in critically ill patients has been controversial. Multiple studies using recombinant EPO have shown conflicting results.50,51 The use of recombinant EPO may result in a small decrease in RBC transfusion for some patients, but it does not show an overall mortality benefit and appears that the risk of thrombotic events likely outweighs the benefits in most critically ill patients.51 The one subset of patients that may benefit from recombinant EPO is patients with multiple trauma,52 but the reason for this is unclear and the practice continues to be controversial.

Transfusion of blood components can be an essential part of the management of critically ill patients. However, despite the growing body of evidence recommending specific transfusion thresholds and more judicious use of blood products, many transfusion practices continue to be rooted in tradition. Transfusion carries with it multiple risks, including an increased risk of infection, MOF, ALI, and mortality for every unit of component administered. Practitioners must have a clear understanding of these risks in order to use blood component therapy safely and effectively.