INTRODUCTION — Blood and coagulation management are key components of anesthetic management for cardiac surgical procedures that require cardiopulmonary bypass (CPB) because of the need for full systemic anticoagulation during CPB and its reversal after weaning from CPB. Other major factors influencing coagulation include hemodilution due to fluid priming of the extracorporeal circuit, fibrinolysis, platelet dysfunction, coagulation factor consumption, and systemic hypothermia that occurs during CPB, as well as blood loss during surgery involving the heart and great vessels [1,2].

This topic discusses reversal of anticoagulation and management of bleeding after cardiac surgery with CPB. Strategies to avoid or minimize blood loss and transfusion of blood products before and during CPB are discussed separately. (See "Blood management and anticoagulation for cardiopulmonary bypass".)

General principles for perioperative blood management are discussed in separate topics. (See "Perioperative blood management: Strategies to minimize transfusions" and "Intraoperative transfusion of blood products in adults".)

REVERSAL OF ANTICOAGULATION — Systemic heparin anticoagulation is reversed with administration of protamine after weaning from CPB [3].

Administration and dosing of protamine

●Timing of initial administration – Neutralization of systemic heparin with protamine is typically initiated after weaning from CPB but before aortic decannulation. This timing allows rapid reinstitution of CPB in the event of a catastrophic protamine reaction [4]. An initial "test dose" of protamine 10 mg is administered to allow early detection of serious hemodynamic changes that may indicate a protamine reaction. If there is no adverse reaction, the remainder of the protamine dose is administered slowly, typically over a 5 to 15 minute period, preferably by infusion [5,6]. Slower infusion may avoid potential vasodilation. (See 'Management of protamine reactions' below.)

Suctioning of blood from the surgical field into the pump reservoir (ie, "cardiotomy suction" or "pump suckers") is discontinued either at the onset of protamine administration or when a small portion (typically less than one-third) of the total initial dose has been administered. It is hypothesized that neutralization of heparin in the blood remaining in the CPB reservoir may cause clot formation, which would prohibit emergency reinstitution of CPB.

●Initial dosing – Calculation of protamine dosing for heparin reversal is ideally based on a point-of-care (POC) titration to existing heparin in the blood [7], or if a POC heparin-protamine titration assay is not available, by administering 0.7 mg protamine per 100 units of the initial pre-bypass heparin dose. After administration of the initial calculated protamine dose, further evaluation for residual heparin effect is accomplished by measuring the activated clotting time (ACT), rechecking the heparin-protamine titration assay, and/or by checking viscoelastic clot strength with POC tests of hemostatic function [8-11] (see "Intraoperative transfusion of blood products in adults", section on 'Overall hemostatic function'). Plasma concentrations of residual heparin can insidiously increase after administration of the initial protamine dose as heparinized blood from the CPB pump is reinfused, and as peripheral tissues gradually release endothelial bound heparin into the circulation as reperfusion occurs. Additional protamine is administered according to heparin-protamine titration assay measurements of residual heparin, or an additional 25 to 100 mg of protamine may be administered if evidence of residual heparin effect is noted on the ACT or with viscoelastic testing. However, protamine overdosing should be avoided [3]. In most cases, there is no reason to exceed 1 mg protamine per 100 units of total administered heparin; in practice, we rarely administer more than 250 mg of protamine.

There are numerous institutional variations in protamine dosing practices to achieve heparin neutralization. Traditional methods relied on fixed protamine dosing based on body weight and/or the amount of heparin administered. A 2013 meta-analysis of these traditional methods versus use of a heparin-protamine titration assay found significantly less mean postoperative blood loss with protamine titration compared with standard (fixed) dosing [12]. Other studies have noted that excess protamine is associated with prolonged ACT, international normalized ratio (INR), and activated partial thromboplastin time (aPTT) values, as well as inhibited platelet function, factor V activation, and excessive bleeding after CPB, particularly if the ratio of protamine to heparin exceeds 2.6 mg protamine per 100 units of heparin [7,8,13-17]. We agree with practice guidelines stating that it is reasonable to limit this ratio to ≤2.6 [7,18].

●Postbypass and postoperative protamine administration – We administer a protamine infusion of 25 to 50 mg/hour over two to four postoperative hours to potentially reduce blood loss by avoiding "heparin rebound" [7,18-20]. Release of heparin into the intravascular space may occur after the initial protamine dose has been metabolized. This "heparin rebound" is particularly likely in patients who received a large dose of heparin due to altered heparin responsiveness [7,18,21] (see "Blood management and anticoagulation for cardiopulmonary bypass", section on 'Heparin resistance'). Also, heparin may accumulate in adipose tissue due to its lipophilicity, with slow eventual release [19,22]. Furthermore, active cooling during CPB causes vasoconstriction in peripheral tissues, especially adipose tissue, leading to a potential depot of heparin for later release after reperfusion.

Administration of a postbypass and postoperative protamine infusion allows optimal neutralization of heparin that may be slowly released from such "depots." Heparin levels measured during these phases usually range from 0.1 to 0.3 international units/mL, which is at the low end of the therapeutic range. Notably, the ACT and even activated partial thromboplastin time (aPTT) are poorly sensitive measurements of residual heparin in this low range; plasma heparin concentrations may be more accurate [23,24].

Management of protamine reactions — Mechanisms responsible for protamine reactions include:

●Vasodilation due to direct and indirect effects of protamine and potential complement activation by heparin-protamine complexes [4,18,25-27]. Such reactions are common, associated with a faster rate of protamine infusion, and are treated with administration of vasopressors as needed (table 1).

●A more severe anaphylactic reaction with vasoplegia is less common, but may occur in individuals who have antibodies to protamine due to previous exposure (eg, protamine-containing insulin) or due to allergy [4,7,18,25,26,28-30]. In such cases, cardiovascular collapse may be unrelated to the rate of protamine administration. These reactions are treated in the same manner as other perioperative anaphylactic reactions (table 2). (See "Perioperative anaphylaxis: Clinical manifestations, etiology, and management", section on 'Initial management'.)

●Acute pulmonary vasoconstriction leading to right ventricular failure is more rare, and may be accompanied by bronchospasm or noncardiogenic pulmonary edema due to thromboxane release from platelets [4,7,18,25-27,29,31-33]. Such severe reactions are likely caused by either IgG antibodies or heparin-protamine complexes that trigger thromboxane release. The differential diagnosis includes blood product administration causing transfusion-related acute lung injury (TRALI), which can produce a similar reaction [34]. Treatment consists of stopping protamine and implementing resuscitative measures. Systemic reheparinization may be necessary to allow immediate reinstitution of CPB [7,31].

When resuscitation from any severe protamine reaction requires heparin readministration for reinstitution of CPB for temporary hemodynamic support, we administer anti-inflammatory and vasoactive agents. Specifically, we administer a systemic steroid dose (eg, methylprednisolone 125 mg), together with the H1 antihistamine diphenhydramine 50 mg and a histamine-2 receptor antagonist (eg, famotidine, cimetidine), as noted in the table (table 2). An infusion of epinephrine should be administered, and norepinephrine or vasopressin should be initiated if necessary to treat refractory hypotension (table 1). In patients with persistent refractory vasodilatory shock, it is reasonable to administer a dose of methylene blue 1 to 2 mg/kg before attempting to wean from CPB [35-37].

For patients with suspected allergic reaction, a blood sample should be collected at or shortly after the event, which may reveal elevations in tryptase. After postoperative recovery patients are referred to an allergy specialist. (See "Perioperative anaphylaxis: Clinical manifestations, etiology, and management", section on 'Laboratory tests at the time of the reaction' and "Perioperative anaphylaxis: Clinical manifestations, etiology, and management", section on 'Referral for allergy evaluation'.)

After stabilization on CPB we consider next steps in consultation with members of the anesthesiology, surgical, and perfusionist teams. There are no clear guidelines regarding the safety of readministering protamine to neutralize the additional systemic dose of heparin after a catastrophic protamine reaction [18,27,30]. In some cases, clinicians have safely readministered protamine without incident [30]. In theory, the initial anaphylactic reaction leads to a post-anaphylaxis refractory period due to temporary depletion of inflammatory mediators [30,34]. However, some clinicians have avoided readministration of protamine, electing instead to reverse anticoagulation with blood products. With this option, excessive bleeding may require transfusion of large volumes of fresh frozen plasma (FFP), platelets, and fibrinogen, as well as red blood cells (RBCs), with the associated adverse effects of massive transfusion [30,38]. (See "Massive blood transfusion".)

In rare cases when noncardiogenic pulmonary edema or respiratory distress syndrome is severe after a protamine reaction, temporary extracorporeal membrane oxygenation may be necessary after weaning from CPB. (See "Intraoperative problems after cardiopulmonary bypass", section on 'Extracorporeal membrane oxygenation'.)

Alternatives to protamine — There are no approved alternatives to protamine for heparin reversal [18]. With a half-life of approximately 60 to 90 minutes, the anticoagulant effect of heparin may persist for four to six hours if not reversed, although blood loss replacement with blood products and/or volume expanders can accelerate the decrease in heparin concentration.

Bivalirudin is an alternative anticoagulant agent used to avoid heparin and protamine in patients with known severe allergy to either agent, or those with heparin-induced thrombocytopenia (HIT) with HIT antibodies present (table 3) (see "Management of heparin-induced thrombocytopenia (HIT) during cardiac or vascular surgery"). Bivalirudin has a half-life of approximately 45 minutes in normothermic patients with normal renal function; thus, all anticoagulant effects typically resolve within approximately two hours of administration of the last dose. If postbypass bleeding persists after administration of bivalirudin, reversal of anticoagulant effects may be achieved using combinations of therapies such as blood product replacement, hemodialysis, or plasmapheresis with plasma exchange [18].

MANAGEMENT OF BLEEDING — Bleeding necessitating transfusion occurs commonly after cardiac surgery with CPB, with transfusion rates widely varying between institutions (10 to 90 percent) [39,40]. Risk factors for perioperative bleeding and blood transfusion during cardiac surgery include advanced age, decreased preoperative red blood cell (RBC) volume (eg, small body size, preoperative anemia), and complex or redo operations [41-44]. Patients who are transfused have worse outcomes than those without transfusions [45].

Causes of excessive bleeding during and after weaning from CPB include inadequate surgical hemostasis, loss of platelets and coagulation factors due to persistent surgical bleeding, effects of hemodilution, hypothermic coagulopathy, presence of residual heparin (see 'Reversal of anticoagulation' above), and coagulopathy due to platelet activation (and consumption) and hyperfibrinolysis induced by the extracorporeal circuit. (See "Blood management and anticoagulation for cardiopulmonary bypass", section on 'Effects of cardiopulmonary bypass on hemostasis'.)

Fastidious surgical hemostasis — Meticulous surgical technique and systematic intraoperative checking of potential surgical sites of bleeding are critically important after CPB. In a meta-analysis of patients requiring mediastinal re-exploration for bleeding or cardiac tamponade after cardiac surgery, a surgical site of bleeding was identified in two-thirds of cases [46]. Excessive bleeding requiring mediastinal re-exploration has been associated with adverse outcomes that include mortality, need for mechanical circulatory support, stroke, acute renal failure, sternal wound infection, and prolonged mechanical ventilation [47].

Maintenance of normothermia — Mild (32 to 35°C) or moderate (28 to 32°C) hypothermia may be induced to provide neurologic and cardiac protection for many patients undergoing CPB, while deep hypothermia to temperatures as low as 16 to 18°C may be employed for selected patients undergoing elective circulatory arrest. (See "Management of cardiopulmonary bypass", section on 'Management during cooling and hypothermia' and "Anesthesia for aortic surgery requiring deep hypothermia", section on 'Effects of deep hypothermia'.)

Hypothermia is associated with coagulopathy due to impairment of platelet aggregation and reduced activity of clotting enzymes [48-50]. This combination of platelet and enzyme impairment typically reduces clot formation and increases perioperative blood loss and the need for transfusion [51-53].

Active warming techniques must be employed to achieve normothermia during the rewarming phase of CPB, then to maintain normothermia during the postbypass and postoperative periods. Further details regarding warming strategies are available in other topics:

●(See "Management of cardiopulmonary bypass", section on 'Management during rewarming and weaning'.)

●(See "Anesthesia for aortic surgery requiring deep hypothermia", section on 'Rewarming strategies'.)

●(See "Perioperative temperature management", section on 'Active warming devices'.)

Use of transfusion algorithms — Similar to practice guidelines of several professional societies, we use a goal-directed protocol or algorithm to guide transfusion decisions, based on measurement of hemoglobin (Hgb) or hematocrit (Hct), as well as assessment of specific abnormalities of hemostasis using standard laboratory tests (algorithm 1), and/or point-of-care (POC) tests of hemostasis [40,41,54-57]. Use of such transfusion protocols and algorithms to guide decision-making in cardiac surgery can avoid or reduce unnecessary transfusions of blood products including RBCs, fresh frozen plasma (FFP), platelets, and cryoprecipitate [40,57-68]. (See "Intraoperative transfusion of blood products in adults", section on 'Use of a transfusion algorithm or guideline'.)

●Hemoglobin or hematocrit levels – We check Hgb or Hct levels following weaning from CPB, and if the patient is actively bleeding, approximately every 30 minutes or more frequently in the profusely bleeding patient. We transfuse RBCs if Hgb level is <7 to 8 g/dL (or Hct <21 to 24 percent) [41,42] but may target a higher Hgb transfusion threshold in patients with uncontrolled hemorrhage. (See 'Red blood cells' below.)

●Standard laboratory coagulation tests – Standard laboratory tests include prothrombin time (PT), activated partial thromboplastin time (aPTT), international normalized ratio (INR), fibrinogen level, and platelet count. Transfusion algorithms generally recommend administration of platelets for a platelet count <50,000 to 100,000, FFP when persistent bleeding is accompanied by a PT/INR or aPTT >1.5 times normal value, or cryoprecipitate or fibrinogen concentrate when fibrinogen levels are <150 to 200 mg/dL. (See 'Fresh frozen plasma' below and 'Platelets' below and 'Cryoprecipitate' below.)

When available, tests of platelet function are performed (table 4). However, testing for platelet dysfunction is unreliable in the actively bleeding patient since most tests need a relatively normal platelet count. Furthermore, most platelet function tests are not accurate following dilutional changes and platelet activation after CPB [69]. (See "Clinical use of coagulation tests" and "Platelet function testing".)

●Point-of-care tests of hemostatic function – Similar to recommendations made in the 2019 Society of Cardiovascular Anesthesiologists Clinical Practice Improvement Advisory for Management of Perioperative Bleeding and Hemostasis in Cardiac Surgical Patients, we employ POC viscoelastic coagulation tests to guide transfusion therapy [40]. Such viscoelastic coagulation tests (eg, thromboelastography [TEG] or an adaptation of TEG known as rotational thromboelastometry [ROTEM] (table 5)) supplement standard laboratory tests of hemostatic function, and may be superior to standard tests and/or clinical judgment by allowing more rapid assessments of coagulopathy and responses to interventions (eg, transfusion of blood products, administration of hemostatic agents). Meta-analyses of randomized trials in cardiac surgical patients have noted that use of transfusion algorithms guided by viscoelastic testing reduces RBC and platelet transfusions compared with standard care (eg, standard laboratory coagulation tests and/or clinical judgment) [70-73]. Observational studies have reported similar results [62,64-66,68,74,75]. However, POC viscoelastic tests are not available in every institution, and conventional laboratory coagulation tests are also recommended as additional or alternative assessment tools [40].

Notably, platelet function defects due to aspirin, dipyridamole, P2Y₁₂ receptor antagonists, or lower levels of reversible GP IIbIIIa antagonists (eg, eptifibatide) may not be detected by these standard viscoelastic tests because thrombin is generated in the TEG or ROTEM sample cups and produce a "false" normal test result despite the presence of clinical coagulopathy in the patient. This is consistent with the responses to thrombin agonists during platelet aggregometry testing, and is important to recognize when interpreting TEG/ROTEM results.

Clinically significant bleeding at any time during the postbypass period should trigger repeat evaluation of standard laboratory tests (algorithm 1), and/or POC tests to determine which algorithm-based therapies are most likely to reverse existing and developing abnormalities.

Further information regarding POC tests of coagulation is available in separate topics:

•(See "Intraoperative transfusion of blood products in adults", section on 'Overall hemostatic function'.)

•(See "Clinical use of coagulation tests", section on 'Point-of-care testing'.)

•(See "Coagulopathy in trauma patients", section on 'Thromboelastography-based transfusion'.)

Transfusion of blood products

Red blood cells — We transfuse RBCs for Hgb <7 to 8 g/dL, similar to professional society practice guidelines for blood conservation during cardiac surgery [3,41,42]. However, transfusion decisions are individualized, taking into account patient-related factors (eg, age, severity of illness, cardiac function, risk for critical end-organ ischemia), laboratory parameters (eg, low mixed venous oxygen saturation [SVO₂]), evidence of myocardial ischemia on electrocardiogram (ECG) or transesophageal echocardiography (TEE), and the clinical setting (massive or active blood loss) [41-43,76]. When blood loss is rapid, immediate transfusion may be life-saving and necessary before quantitative laboratory assessment of Hgb can be obtained, based on the rate of bleeding, expected volume of ongoing bleeding, and the preoperative red cell mass [41,56]. When RBC transfusion is necessary, leukocyte-reduced blood is preferred. (See "Intraoperative transfusion of blood products in adults", section on 'Red blood cells' and "Indications and hemoglobin thresholds for red blood cell transfusion in the adult", section on 'Cardiac surgery' and "Leukoreduction to prevent complications of blood transfusion".)

When transfusion is necessary, available salvaged blood is returned first, followed by reinfusion of blood units harvested via normovolemic hemodilution, then allogeneic RBCs [40]. (See "Surgical blood conservation: Blood salvage" and "Blood management and anticoagulation for cardiopulmonary bypass", section on 'Acute normovolemic hemodilution'.)

The optimal Hgb level during and after cardiac surgery is not known. In a multicenter randomized trial conducted in more than 5000 cardiac surgical patients, the primary composite outcome that included death, myocardial infarction, renal failure, and stroke was similar with use of a restrictive transfusion strategy (Hgb trigger set at <7.5 g/dL) compared with use of a liberal transfusion strategy (Hgb trigger set at <9.5 g/dL) during the intraoperative and postoperative periods [77]. In the restrictive group, fewer patients received any RBC transfusion (52 versus 73 percent; odds ratio [OR] 0.41, 95% CI 0.37-0.47), and fewer total RBC units were transfused (two versus three units; rate ratio 0.85. 95% CI 0.82-0.88). At follow-up six months later, the primary composite outcome remained similar (17.4 percent in the restrictive group versus 17.1 percent in the liberal group; OR 1.02, 95% CI 0.87-1.18), and mortality did not differ [78]. A meta-analysis of randomized trials investigating restrictive versus liberal strategies for RBC transfusion in cardiac surgical patients noted similar outcomes for mortality, myocardial infarction, stroke, renal failure, or infection with either strategy [79]. Also, fewer RBC transfusions were reported with a restrictive transfusion strategy in some trials [45,80,81]. However, in one meta-analysis of randomized trials that included both noncardiac and cardiac surgical patients, a trend toward lower mortality was noted if a liberal rather than a restrictive transfusion threshold was employed [82]. This isolated, non-significant finding is insufficient to justify liberal transfusion in the absence of a clinical indication.

In observational studies of cardiac surgical patients, an association between RBC transfusions and adverse postoperative outcomes has been noted (eg, mortality, renal failure, stroke), but these findings are not consistent, nor apparent in the prospective studies discussed above [83,84]. Massive transfusions may pose a greater risk [38,85]. Further discussion is available in a separate topic. (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Blood transfusion'.)

Fresh frozen plasma — We use goal-directed protocols or algorithms to guide transfusion decisions based on specific abnormalities of hemostasis (eg, PT, aPTT, or INR >1.5 times normal value in the setting of ongoing bleeding) [41,86] (see 'Use of transfusion algorithms' above). Transfusion of FFP has many of the risks associated with RBC transfusion, and is not transfused in the absence of significant laboratory evidence of coagulopathy and active bleeding [3,40,86].

Transfusion of FFP is also included in massive transfusion protocols, usually in a 1:1 ratio with transfusion of RBC units, which is relevant to a massively bleeding cardiac surgery patient [87]. (See "Intraoperative transfusion of blood products in adults", section on 'Plasma'.)

Platelets — Transfusion of platelets is reserved for patients with platelet count <100,000/mm³ (or platelet dysfunction due to residual anti-platelet drug effect) if there is ongoing clinically significant microvascular bleeding [40]. Patients who are taking antiplatelet medication, particularly P2Y₁₂ receptor inhibitors such as clopidogrel, prasugrel, and ticagrelor, are at higher risk for such bleeding due to platelet dysfunction. (See "Intraoperative transfusion of blood products in adults", section on 'Platelets' and "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Antiplatelet agents'.)

Thrombocytopenia is common in the immediate postbypass period due to a combination of hemodilution, platelet loss due to persistent surgical bleeding, platelet adherence to the CPB circuit surface, consumption due to coagulation activation, and accelerated clearance caused by thrombin-mediated activation. Rarely, thrombocytopenia is a manifestation of acute disseminated intravascular coagulation. (See "Blood management and anticoagulation for cardiopulmonary bypass", section on 'Effects of cardiopulmonary bypass on hemostasis' and "Evaluation and management of disseminated intravascular coagulation (DIC) in adults", section on 'Acute DIC'.)

Management of thrombocytopenia and thrombosis due to heparin-induced thrombocytopenia (HIT) in the postoperative period is discussed separately. (See "Management of heparin-induced thrombocytopenia (HIT) during cardiac or vascular surgery".)

Cryoprecipitate — Cryoprecipitate can be used to treat hypofibrinogenemia (eg, <100 mg/dL) in a patient with intractable bleeding and coagulopathy when fibrinogen concentrate is not available (see 'Fibrinogen concentrate versus cryoprecipitate' below). One unit of cryoprecipitate contains all of the fibrinogen present in one unit of whole blood (approximately 200 to 400 mg) in a volume of 10 to 20 mL (table 6). Each unit raises the plasma fibrinogen concentration by approximately 7 to 10 mg/dL. For adult patients, cryoprecipitate is typically prepared as a pooled product combining individual cryoprecipitate units derived from 5 to 10 blood donors in a final volume of approximately 50 to 200 mL. Due to the low volume of antibodies in each unit, cryoprecipitate does not need to be ABO matched in adults. (See "Clinical use of Cryoprecipitate".)

Compared with fibrinogen concentrate (see 'Fibrinogen concentrate versus cryoprecipitate' below), cryoprecipitate is more readily available with lower cost in the United States, and is used more commonly in both the United States and the United Kingdom. Also, bleeding and coagulopathy are often multifactorial in the immediate postbypass period. For this reason, many clinicians prefer cryoprecipitate to treat significant postbypass coagulopathy since each unit is derived from one unit of FFP and thus replaces many hemostatic factors (eg, fibrinogen, factor VIII, factor XIII, von Willebrand factor [vWF], fibronectin) (table 7). However, cryoprecipitate is not available on mainland Europe.

Dosing with a 5 unit bag of cryoprecipitate will increase fibrinogen levels by approximately 35 to 50 mg/dL in a 70 kg adult, although this increase may be less if the patient is actively hemorrhaging [88]. As with fibrinogen concentrate, the plasma fibrinogen level is monitored, and repeat doses are administered to maintain the level above the appropriate threshold. (See "Intraoperative transfusion of blood products in adults", section on 'Cryoprecipitate'.)

Risks of cryoprecipitate transfusion are similar to those for FFP, although transfusion-associated circulatory overload (TACO) is less likely with cryoprecipitate due to lower total transfused volume. (See "Clinical use of Cryoprecipitate", section on 'Risks and adverse events'.)

Use of hemostatic agents — Hemostatic agents may be used to treat excessive bleeding in the postbypass period.

Antifibrinolytic agents — Timing for discontinuing administration of a prophylactic antifibrinolytic agent (eg, epsilon-aminocaproic acid [EACA] or tranexamic acid [TXA]) in the postbypass or postoperative period varies among centers. However, continuing administration is reasonable for patients with persistent bleeding [89,90]. (See "Blood management and anticoagulation for cardiopulmonary bypass", section on 'Antifibrinolytic administration'.)

Fibrinogen concentrate versus cryoprecipitate — We administer a fibrinogen concentrate, if available, to treat hypofibrinogenemia (eg, <100 mg/dL) in cardiac surgical patients with clinically significant bleeding and coagulopathy [40,91-99]. Fibrinogen concentrate can be used for treatment of hypofibrinogenemia rather than using cryoprecipitate or FFP (see 'Cryoprecipitate' above and 'Fresh frozen plasma' above). Fibrinogen concentrate is a pasteurized and pooled product that has a small volume and offers less risk of infection transmission and immunological complications of blood transfusion [99]. It is also a pharmaceutical product that can be stored and reconstituted at the point of care.

Fibrinogen concentrates (eg, RiaSTAP, Haemocomplettan, FIBRYGA) are prepared from pooled human plasma and are available as lyophilized powders (approximately 1 g [1000 mg]/vial) that are reconstituted in a small volume. Dose is calculated according to the target fibrinogen concentration; administration of 70 mg/kg increases fibrinogen concentration by approximately 100 mg/dL [100]. Dosing is typically monitored with measurements of plasma fibrinogen levels as well as POC testing to avoid hypercoagulability. (See "Intraoperative transfusion of blood products in adults", section on 'Point-of-care tests' and "Disorders of fibrinogen", section on 'Fibrinogen concentrate: Dosing and monitoring'.)  

Meta-analyses have found that administration of fibrinogen concentrate reduces transfusions in many types of cardiac surgical procedures, but effects on other outcomes such as survival are unclear:

●A 2019 trial randomly assigned 735 cardiac surgical patients with clinically significant bleeding and documented hypofibrinogenemia (fibrinogen level <150 to 200 mg/dL) to receive 4 g of fibrinogen concentrate or 10 units of cryoprecipitate after CPB [101]. Both groups received a similar number of units of allogeneic blood component transfusions including RBCs, FFP, and platelets. The number of units of transfused allogeneic blood components including RBCs, FFP, and platelets was similar for patients receiving either treatment. Differences between the groups were not significant for other outcomes including mortality (9.4 percent in the fibrinogen concentrate group versus 7.4 percent in the cryoprecipitate group; hazard ratio [HR] 1.28; 95% CI 0.77-2.12) and thromboembolic complications (7.0 percent in the fibrinogen concentrate group versus 9.6 percent in the cryoprecipitate group; OR 0.70; 95% CI 0.42-1.20).

●A 2018 meta-analysis of randomized trials (exclusively cardiac surgical patients) found that fibrinogen concentrate administered either prophylactically before CPB or during bleeding after CPB reduced the risk of RBC transfusion compared with placebo or other hemostatic treatments (risk ratio 0.64, 95% CI 0.49-0.83), but mortality was not reduced (597 patients; eight trials) [98]. Conclusions were limited by the small number of clinical events within each trial.

●In two randomized trials in patients undergoing high-risk complex cardiac surgical procedures, prophylactic use of fibrinogen concentrate was not effective in reducing blood loss or transfusion of allogeneic blood products compared with placebo [102,103].

Administration of fibrinogen concentrate to treat severe perioperative bleeding is more common in Europe than in the United States, in part because cryoprecipitate is not typically available in Europe. The European Society of Anesthesia guidelines suggest use of fibrinogen concentrate to maintain a target fibrinogen concentration >150 to 200 mg/dL [104,105]. Similarly, the Hemostasis and Transfusion Scientific subcommittee of the European Association of Cardiothoracic Anaesthesiology suggests administration of fibrinogen concentrate in cardiac surgical patients with microvascular bleeding after CPB to maintain physiologic fibrinogen activity based on viscoelastic coagulation testing [99]. Reduced serum concentrations of fibrinogen may predict bleeding since it is the first substrate component of the coagulation cascade to be consumed below critical levels in hemorrhaging patients [106]. Proponents argue that the baseline concentration of fibrinogen is relatively low, and there are no fibrinogen stores to be mobilized [91,107]. Thus, fibrinogen is the first coagulation protein to become critically low during intraoperative bleeding, and it may be the most critical coagulation component required for local hemostasis [88,106]. However, guidelines do not suggest aiming for supranormal fibrinogen levels [99]. Notably, administration of fibrinogen concentrate as a single agent will only raise the fibrinogen level and will not address other coagulation factor deficiencies. For this reason, efficacy compared with placebo has not been shown when fibrinogen concentrate is administered prophylactically in patients with expected significant intraoperative bleeding during complex cardiac surgical procedures [102,103]. While administration of fibrinogen concentrate will not always resolve coagulopathic bleeding, we aggressively correct acquired hypofibrinogenemia in patients with clinically significant bleeding as a component of a multifactorial transfusion algorithm [58]. Some clinicians prefer cryoprecipitate in this setting since each unit contains many hemostatic factors (eg, fibrinogen, factor VIII, factor XIII, von Willebrand factor [vWF], fibronectin), as noted above. (See 'Cryoprecipitate' above.)

Although thromboembolic complications can occur, particularly in patients with a thrombotic fibrinogen variant (or pregnancy), most studies have not reported increased risk of thrombotic events [96,98,99]. Monitoring to avoid overcorrection of fibrinogen deficiency minimizes this risk.

Prothrombin complex concentrate (PCC)

●PCCs – A 4-factor unactivated prothrombin complex concentrate (PCC) is typically used for emergency cardiac surgery in patients chronically taking warfarin or another vitamin K antagonist [40,108]. All PCCs contain factors II, IX and X. Those that do not contain appreciable factor VII are known as 3-factor PCCs, while those containing factor VII are labeled as 4-factor PCCs (table 8). Advantages of PCCs over FFP include rapid administration in a small volume, resulting in more rapid reversal of the anticoagulant effect and avoidance of volume overload and transfusion reactions [91,93,108-113]. (See "Plasma derivatives and recombinant DNA-produced coagulation factors", section on 'PCCs'.)

For patients with intractable coagulopathy and diffuse bleeding or for patients intolerant of high FFP transfusion volume, off-label intraoperative use of unactivated 4-factor or 3-factor PCC products after cardiac surgery has been evaluated in observational studies [40,91,109,110,114-119]. However, other primary causes such as thrombocytopenia or low levels of fibrinogen due to hemodilution, platelet dysfunction, or surgical sources of bleeding should be sought and treated before considering a PCC product for such intractable coagulopathy [109,120]. Also, consider that PCC does contain antithrombin III, while FFP does. Finally, risks of PCC administration include thromboembolic events, which may be more likely with repeat or excessive dosing of both 4-factor and 3-factor PCCs, and may extend well into the postoperative period [121].

Data regarding intraoperative safety of PCC products in cardiac surgical patients are limited [104,109,119,122,123]:

•A 2019 meta-analysis that included 861 patients in four nonrandomized studies in cardiac surgical patients noted a reduced risk of RBC transfusion with administration of perioperative PCC compared with administration of FFP (OR 2.22, 95% CI 1.45-3.40) [119]. There were no differences in the risk of in-hospital mortality, stroke, or acute kidney injury between patients receiving PCC versus FFP (although there was a trend toward a higher risk of renal replacement therapy in the PCC group).

•A subsequently published retrospective study analyzed 114 consecutive patients who received a low dose of 3-factor PCC (ie, 15.8 ± 7.1 international units/kg) for refractory bleeding during complex cardiac surgical procedure as part of a transfusion algorithm guided by standard and POC laboratory testing (see 'Use of transfusion algorithms' above) [123]. Administration of the 3-factor PCC reduced transfusions of blood products compared with the period before PCC administration. However, four patients receiving PCC suffered an ischemic stroke within 30 days of surgery (3.8 percent), while seven had a venous thromboembolic event (6.5 percent) [123].

As more patients are entered into registries for 4-factor and 3-factor PCC product use, risk assessment for such dose-related adverse events will be more certain.

●Activated PCCs – In contrast with unactivated PCCs, activated PCC products such as factor eight inhibitor bypassing activity (FEIBA) contain activated factor VII (table 8) [109]. These agents have a greater prothrombotic risk compared with unactivated PCC products, and are only rarely used. (See "Plasma derivatives and recombinant DNA-produced coagulation factors", section on 'PCCs'.)

Recombinant activated factor VII (rFVIIa) — Recombinant activated factor VII (rFVIIa) is licensed for prevention of surgical bleeding in patients with hemophilia. In rare instances when intractable life-threatening bleeding persists after cardiac surgery with CPB, rFVIIa has been administered and may achieve bleeding cessation and reduce transfusion requirements in that setting [40,124-126]. However, high thromboembolism rates >20 percent (including myocardial infarction) with mortality >30 percent have been described in retrospective evaluations of registries for refractory bleeding cases [127-129]. (See "Recombinant factor VIIa: Administration and adverse effects", section on 'Off-label uses' and "Perioperative blood management: Strategies to minimize transfusions", section on 'Recombinant activated factor VII'.)

Although the optimal dose of rFVIIa for off-label use is unknown, relatively low doses may achieve bleeding cessation and are preferred in most cardiac surgical patients with life-threatening bleeding [40]. One study reported that a median dose of 13.3 mcg/kg resulted in low overall blood transfusions in this setting [125]. A cautious dosing strategy is necessary in the rare instances when rFVIIa is employed. For example, small incremental doses of 10 mcg/kg can be administered approximately every 15 minutes, while recognizing that thromboembolic complications will likely increase with dose escalation, or in the presence of stagnant flow or devices such as extracorporeal membrane oxygenation (ECMO). Stocking the smaller 1 mg vials of rFVIIa facilitates this incremental approach.

Similar to off-label use of PCCs, other primary causes of coagulopathy should always be sought and treated before administering rFVIIa. Failure to treat the primary coagulation defect increases the likelihood of an inadequate response to the initial dose of rFVIIa [127,128,130], and may encourage use of higher doses that are more likely to cause thrombosis [129,131]. (See "Recombinant factor VIIa: Administration and adverse effects", section on 'General approach to administration'.)

Desmopressin (DDAVP) — We administer intravenous (IV) desmopressin (DDAVP) 0.3 mcg/kg when persistent microvascular bleeding is evident after CPB in selected patients with acquired platelet defects due to uremia, or to those with acquired von Willebrand syndrome (eg, due to chronic aortic stenosis or presence of a left ventricular assist device) [91,132-135]. DDAVP should be infused slowly over 15 to 30 minutes to avoid vasodilation. (See "Platelet dysfunction in uremia", section on 'Desmopressin (DDAVP)' and "Acquired von Willebrand syndrome", section on 'Treatment of acute bleeding'.)

Very limited data suggest that IV DDAVP may be beneficial to reduce blood loss in some cardiac surgical patients with intractable microvascular bleeding due to platelet dysfunction that may be caused by hypothermia, acidosis, aspirin use, and/or the effects of CPB [40,136-139]. However, studies are not consistent [140,141]. Two 2017 meta-analyses of the efficacy of DDAVP in patients undergoing cardiac or noncardiac surgery noted only small reductions in perioperative blood loss and volume of RBC transfusions compared with placebo, but quality of evidence was low [138,139]. Thus, generalized perioperative use in cardiac surgical patients is not warranted [40,54,91].

Possible adverse side effects of DDAVP include hypertension, hypotension, and flushing, as well as fluid overload, hyponatremia (which may cause seizures if close attention to free water restriction is not used), and rare thrombotic events (table 9). Tachyphylaxis occurs after the initial dose and during the initial three to five days of administration.

EARLY POSTOPERATIVE MANAGEMENT

Management of bleeding and coagulopathy — Close monitoring for bleeding continues in the postoperative period. Medical causes of bleeding should be investigated and treated [142].

Return to the operating room for surgical re-exploration and intervention may be necessary based on the rate, and presumed location of bleeding, as well as potential for a surgical cause [143,144]. (See "Perioperative blood management: Strategies to minimize transfusions", section on 'Close monitoring for bleeding'.)

Management of anemia — Postoperative anemia is common due to exacerbation of pre-existing anemia, blood loss during surgery, and excessive postoperative phlebotomies [86]. Management strategies are discussed separately (algorithm 2). (See "Perioperative blood management: Strategies to minimize transfusions", section on 'Management of postoperative anemia'.)

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" and "Society guideline links: Management of cardiopulmonary bypass".)

SUMMARY AND RECOMMENDATIONS

●Systemic heparin anticoagulation is reversed with administration of protamine. (See 'Administration and dosing of protamine' above.)

•Timing of initial administration – Neutralization of heparin is initiated after weaning from cardiopulmonary bypass (CPB), but before aortic decannulation. An initial "test dose" of protamine 10 mg is administered to allow early detection of serious hemodynamic changes indicative of a protamine reaction. The remainder of the protamine dose is administered slowly, typically over a 5 to 15 minute period.

•Initial dosing – Calculation of protamine dosing for heparin reversal is based on a point-of-care (POC) titration to existing heparin in the blood, or by administering 0.7 mg protamine per 100 units of the initial pre-bypass heparin dose. Additional protamine is administered according to assay measurements of residual heparin, or an additional 25 to 100 mg of protamine may be administered if evidence of residual heparin effect is noted on the activated clotting time (ACT) measurement or with viscoelastic testing. In most cases, there is no reason to exceed 1 mg protamine per 100 units of total administered heparin; in practice, we rarely administer more than 250 mg of protamine.

•Postbypass and postoperative protamine administration – We administer a protamine infusion of 25 to 50 mg/hour over two to four postoperative hours to avoid "heparin rebound." Release of heparin from the fatty tissues into the intravascular space (ie, "heparin rebound") may occur after the initial protamine dose has been metabolized in patients who received a large dose of heparin.

●Vasodilation due to direct and indirect effects of protamine commonly occur, particularly if administration is rapid. Such reactions are treated with administration of vasopressors as needed (table 1). More severe reactions (ie, anaphylaxis and acute pulmonary vasoconstriction) are uncommon; treatment is described in the table (table 2). (See 'Management of protamine reactions' above.)

●Active warming techniques to achieve and maintain normothermia are employed to minimize coagulopathy and blood loss. (See 'Maintenance of normothermia' above.)

●Causes of excessive bleeding during and after weaning from CPB include inadequate surgical hemostasis, loss of platelets and coagulation factors due to persistent surgical bleeding, effects of hemodilution, presence of residual heparin, hypothermic coagulopathy, and coagulopathy due to hyperfibrinolysis and platelet activation and consumption induced by the extracorporeal circuit. Meticulous surgical techniques and systematic intraoperative checking of potential surgical sites of bleeding is critically important in the postbypass period. (See 'Management of bleeding' above and 'Fastidious surgical hemostasis' above and 'Maintenance of normothermia' above.)

●We use a goal-directed protocol or algorithm to guide transfusion decisions, based on measurement of hemoglobin (Hgb) or hematocrit (Hct) as well as assessment of specific abnormalities of hemostasis using standard laboratory tests (algorithm 1), and/or point-of care tests of hemostasis, similar to the practice guidelines of several professional societies. Use of such transfusion protocols and algorithms to guide decision-making in cardiac surgery can avoid or reduce unnecessary transfusions of blood products. (See 'Use of transfusion algorithms' above.)

●Transfusion of blood components is based on the following parameters (see 'Transfusion of blood products' above):

•Red blood cells (RBCs) – We typically transfuse RBCs for Hgb <7 to 8 g/dL (or Hct < 21 to 24 percent). However, transfusion decisions are individualized, taking into account patient-related factors (eg, age, severity of illness, cardiac function, risk for critical end-organ ischemia), the clinical setting (massive or active blood loss), laboratory parameters (eg, low mixed venous oxygen saturation), or evidence of myocardial ischemia. (See 'Red blood cells' above.)

•Fresh frozen plasma (FFP) – FFP is transfused when persistent bleeding is accompanied by a prothrombin time (PT), activated partial thromboplastin time (aPTT), or international normalized ratio (INR) >1.5 times normal value. (See 'Fresh frozen plasma' above.)

•Platelets – Transfusion of platelets is reserved for patients with platelet count <100,000/mm³ and ongoing clinically significant microvascular bleeding. (See 'Platelets' above.)

•Cryoprecipitate – Cryoprecipitate can be used to treat hypofibrinogenemia (eg, <100 mg/dL) in a patient with intractable bleeding and coagulopathy after CPB, if fibrinogen concentrate is not available. Some clinicians prefer cryoprecipitate in this setting since each unit is derived from one unit of FFP and is a concentrate of many hemostatic factors (eg, fibrinogen, factor VIII, factor XIII, von Willebrand factor [vWF], fibronectin). (See 'Cryoprecipitate' above and 'Fibrinogen concentrate versus cryoprecipitate' above.)

●Hemostatic agents may be used to treat intractable bleeding and coagulopathy after CPB in selected patients:

•Fibrinogen concentrate versus cryoprecipitate – When hypofibrinogenemia is present in a patient with clinically significant bleeding and coagulopathy, a fibrinogen concentrate can be administered. Fibrinogen concentrate or cryoprecipitate is preferred rather than FFP to treat hypofibrinogenemia to avoid volume overload. Fibrinogen concentrate also has little risk of infection transmission or immunological complications of blood transfusion. However, administration of fibrinogen concentrate will only raise the fibrinogen level and will not address other coagulation factor deficiencies. (See 'Fibrinogen concentrate versus cryoprecipitate' above and 'Cryoprecipitate' above.)

•Prothrombin complex concentrate (PCC) – For patients with intractable coagulopathy and diffuse bleeding after CPB, or for patients intolerant of high FFP transfusion volume, off-label intraoperative use of unactivated 4-factor or 3-factor PCC products may be considered. However, other primary causes of bleeding such as thrombocytopenia, hypofibrinogenemia due to hemodilution, platelet dysfunction, or surgical sources of bleeding should be sought and treated before considering a PCC product. Risks include thromboembolic events, which may be more likely with repeat or excessive dosing of both 4-factor and 3-factor PCCs, and may extend well into the postoperative period. (See 'Prothrombin complex concentrate (PCC)' above.)

•Recombinant activated factor VII (rFVIIa) – In rare instances when intractable life-threatening bleeding persists after cardiac surgery, rFVIIa has been administered to achieve bleeding cessation. However, retrospective evaluations have noted high thromboembolism rates after administration. Thus, other primary causes of coagulopathy should always be sought and treated before considering rFVIIa. (See 'Recombinant activated factor VII (rFVIIa)' above.)

•Desmopressin (DDAVP) – We administer intravenous (IV) DDAVP 0.3 mcg/kg when persistent microvascular bleeding is evident after CPB in selected patients with acquired platelet defects due to uremia or those with acquired von Willebrand disease (eg, due to chronic aortic stenosis or presence of a left ventricular assist device). (See 'Desmopressin (DDAVP)' above.)