Concurrent with the development of cardiac surgery in the 1950s as a means of correcting congenital heart defects came the need for large-volume blood transfusions. In the 1960s and 1970s, the introduction of valve prostheses and direct grafting of coronary arteries also made the repair of acquired heart diseases a possibility. These landmarks, along with the early liberal use of allogeneic blood transfusion therapy, led to rapid growth of our field.

Historically, open-heart surgery has been associated with large transfusion requirements. Some reports suggest that up to 70% of this patient population requires blood transfusions, resulting in an average of 2 to 4 donor exposures per patient.^1,2 It has been reported that cardiac surgery consumes about 20% of the available blood supply in the United States, with similar figures observed worldwide.³ These relatively high rates of transfusions are mostly attributable to cardiopulmonary-bypass-induced phenomena: coagulopathy, platelet dysfunction, and red cell hemolysis all occur to varying degrees as a result of the cardiopulmonary bypass (CPB) circuit.^4-6 Other mechanisms for bleeding include inherited and acquired disorders (including platelet dysfunctions, coagulation factor deficiencies, and derangements leading to excessive fibrinolysis) of numerous etiologies.⁷ While common transfusion reactions such as urticaria and fever are easily managed and largely benign, rarer complications like transfusion-related acute lung injury (TRALI) pose serious risks to patients (eg, the mortality rate for TRALI in critically ill populations ranges from 35–58%^8,9) and provide further impetus for the avoidance of blood product transfusions when at all possible.

Although life-threatening hemorrhage is an obvious absolute indication for blood products, many transfusions are also given to improve oxygen-carrying capacity and to avoid or reverse end-organ ischemia. Despite the potential benefits of transfusing blood to maintain end-organ oxygenation, there is a surprising lack of evidence to support the liberal use of blood transfusions in cardiac surgery. In fact, there is a growing literature base demonstrating that transfusions are associated with an increased risk of both morbidity and mortality.^10-12 For example, red blood cell (RBC) transfusions have been associated with longer intensive care unit lengths of stay as well as with worse short- and long-term survival.^13-15 A large prospective study identified RBC transfusion as the strongest independent predictor of all-cause morbidity and mortality following isolated coronary artery bypass grafting (CABG) and that each unit of blood transfused posed an additive risk for adverse outcomes.¹⁶

Concerns over the negative consequences of blood transfusions are not new. As cardiac surgery grew as a discipline in the 1970s, investigators concomitantly noted an increasing incidence of transfusion-transmitted hepatitis; this public health concern first alerted patients and physicians to the concept of blood conservation. The emergence of the human immunodeficiency virus (HIV) a decade later further heightened interest in this area. Considering (1) the increased awareness of blood-borne infectious diseases, (2) the ever-present shortage of available blood donors, (3) the costs of blood products to both patients and institutions, (4) the needs of special populations like Jehovah’s Witnesses, and (5) the inherent risks from transfusions, a greater effort was made to perform open-heart procedures without blood transfusions, even in high-risk patients. Additionally, randomized trials showing that lower hemoglobin thresholds were tolerated in both critically ill and cardiac surgical patients further contributed to decreasing transfusions and to the acceptance of some degree of anemia in our patients.^17-19

As we will discuss, advances in preoperative screening and patient optimization; improvements in surgical techniques and shorter operative times; the use of perioperative pharmacologics and other technologies designed to curtail blood loss; and the aforementioned tolerance of lower hematocrits, especially on bypass, have allowed for extensive procedures to be routinely performed without significant blood loss and with fewer transfusions.

Comprehensive blood conservation programs that combine such elements (analogous to the use of care bundles in the critical care setting) are likely the most effective strategy in decreasing patients’ exposure to allogeneic blood. In 1991, Ovrum and colleagues established the effectiveness of the simple “core” approach to blood conservation in a cohort of 121 consecutive elective CABG patients –the authors achieved a transfusion rate of 4.1% and required only 0.06 RBC units per patient.²⁰ Later that year, the same group applied these principles to a larger group of 500 elective CABG patients, obtaining similar low rates of transfusion (2.4% of patients) and demonstrating that an impressive 96% of patients received no allogeneic blood products; the authors concluded their multimodal six-step blood conservation program was simple, safe, and cost effective.²¹ More recently, Van der Linden and colleagues developed a blood conservation program that employed a standardized transfusion policy and algorithm-driven protocols aimed at minimizing perioperative blood loss.²² Using this strategy, the group reported a 53% decrease in RBC utilization and a 46% decrease in the number of patients receiving any blood products, without a significant difference in postoperative hemoglobin; moreover, their strategy was also shown to be safe and cost effective.²² It should be noted that the success of such multidisciplinary programs is not limited to large teaching institutions, as similar results have been shown in the community setting.²³

Identification of Patients at Risk of Bleeding or of Requiring Transfusion

An important first step in mitigating bleeding and transfusion requirements is to identify those patients most at risk for blood loss and/or for needing blood product replacement. Five factors have been identified that are consistently associated with increased rates of RBC transfusions in cardiac surgery: low preoperative hemoglobin and/or hematocrit, advanced patient age, female sex, renal insufficiency, and surgery urgency.²⁴ Additional risk factors include a personal and/or family history of any excessive bleeding or the presence of a documented bleeding disorder; preoperative antiplatelet or anticoagulation therapy (eg, aspirin, clopidogrel, warfarin); insulin-dependent diabetes mellitus; decreased left ventricular function; anticipated long CPB time; and type of surgery (with complex valvular and aortic surgeries conferring the greatest risks).²⁵ Current Society of Thoracic Surgery/Society of Cardiovascular Anesthesiologists (STS/SCA) blood conservation guidelines recommend obtaining preoperative hematocrits and platelet counts to aid risk prediction, as abnormalities in these variables are amenable to intervention; preoperative bleeding time may also be determined in high-risk patients, especially those on preoperative antiplatelet agents.²⁶ Of note, preoperative screening of the intrinsic coagulation system is not recommended unless there is a clinical history of a prior bleeding diathesis.²⁶

Awareness of the above-listed risk factors is important as the recognition of specific risks should guide subsequent preoperative patient management with the intent of optimizing a given patient’s risk profile. We will in turn discuss three such preoperative management issues: medication cessation, increasing preoperative RBC mass, and preoperative autologous donation.

Managing Medications Contributing to Bleeding and Transfusion Risks

The regular, preoperative use of antiplatelet medications like aspirin and clopidogrel have been associated with increased perioperative blood loss as well as with the need for blood products in cardiac surgery patients—as such, a thorough understanding of current guidelines pertaining to their use prior to cardiac surgery is requisite. With the advent of newer anticoagulants (eg, direct thrombin inhibitors, direct factor Xa inhibitors), insight into what to do about these drugs must be assimilated with knowledge of older, well-established guidelines on the cessation of traditional vitamin K antagonists. Table 14-1 lists some of the more commonly encountered of these drugs.

ASPIRIN

Aspirin irreversibly inhibits cyclooxygenase-1 and -2, leading to decreased formation of thromboxane A₂ and ultimately to inhibited platelet aggregation. The ability of aspirin to irreversibly induce this qualitative platelet defect has led to its widespread use as a thromboprophylaxing agent in many cardiovascular disease states (eg, coronary artery disease, carotid artery stenosis, after prosthetic valve insertion, or CABG); as such, aspirin is a very frequently encountered drug in our patient population. Current guidelines state that it is reasonable to discontinue aspirin prior to cardiac surgery only in purely elective patients not having acute coronary syndromes.²⁶

ADENOSINE DIPHOSPHATE RECEPTOR INHIBITORS

These antiplatelet drugs (clopidogrel, prasugrel, ticagrelor, and ticlopidine) inhibit the P2Y₁₂ subtype of platelet adenosine diphosphate (ADP) receptors, thus causing irreversible platelet inhibition (of note, ticagrelor is an allosteric antagonist, making its ADP blockage reversible). Members of this class of drugs are frequently used in combination with other antiplatelet agents like aspirin as part of dual antiplatelet therapies for acute coronary syndromes and for thromboprophylaxis in those with stents and/or cerebrovascular disease. ADP receptor inhibitors are felt to confer greater bleeding and transfusion risks than aspirin; as such, guidelines recommend discontinuation of these agents as few as 3 days prior to cardiac surgery (with specific timing determined by a given drug’s half-life of elimination).²⁶ Point-of-care (POC, discussed later) testing assessing platelet responsiveness to clopidogrel, specifically, may be used to identify those nonresponders who are candidates for early operative coronary revascularization and who may thus not require a preoperative cessation period.²⁶

GLYCOPROTEIN IIb/IIIa INHIBITORS

Members of this drug class (eg, abciximab, eptifibatide, tirofiban) prevent platelet aggregation via inhibition of glycoprotein (GP) IIb/IIIa receptors on the surface of platelets; these agents are frequently used during percutaneous coronary interventions and in the treatment of acute coronary syndromes. Like the other previously mentioned high-intensity antiplatelet drugs, GP IIb/IIIa inhibitors are associated with increased bleeding after cardiac operations; as such, these medications should be stopped prior to surgery in order to decrease minor and major bleeding events.²⁶ Exact timing again depends on the half-life of each agent in question.

It should be noted that unfractionated heparin is the notable exception to the cessation recommendations pertaining to the high-intensity antithrombotic drugs outlined by the STS/SCA guidelines: unfractionated heparin is the only agent which can either be discontinued shortly before operation or not at all.²⁶

VITAMIN K ANTAGONISTS

Vitamin K antagonists are anticoagulants that reduce hepatic production of coagulation factors II, VII, IX, and X as well as of proteins C and S—all of which depend on vitamin K for their synthesis. Warfarin, the most widely encountered of these drugs, is used both in the prophylaxis and treatment of thromboembolic disorders (eg, venous or pulmonary clots, prosthetic valve thrombosis); warfarin is similarly used in atrial fibrillation and can serve as an adjunct to reduce systemic embolic risks after myocardial infarction. According to recent guidelines produced by the European Association for Cardio-Thoracic Surgery, patients on warfarin prior to cardiac surgery should be managed in a similar manner to those undergoing major noncardiac surgery.²⁷ That is, warfarin should be discontinued 2 to 4 days before surgery and patients at higher risk of thrombosis should be bridged with intravenous heparin once the international normalized ratio becomes subtherapeutic.²⁷

NOVEL ORAL ANTICOAGULANTS: DIRECT THROMBIN INHIBITORS AND DIRECT FACTOR XA INHIBITORS

Although novel oral anticoagulants (NOACs) belonging to the direct thrombin inhibitor (eg, dabigatran) and direct factor Xa inhibitor (rivaroxaban, apixaban, edoxaban) classes are currently not indicated for patients with mechanical valves, their ease of use and their lack of a monitoring requirement (as compared to warfarin) have increased the popularity of these drugs for patients requiring long-term anticoagulation for nonvalvular reasons. Current guidelines regarding the perioperative use of NOACs recommend these drugs be discontinued 2 to 5 days prior to procedures with a high risk of bleeding, including major abdominal, cardiovascular, and thoracic operations.^28,29 Since the diminution of the anticoagulant effects of NOACs is predictable after their cessation, bridging is typically not required after NOACs are stopped.

The management of patients on NOACs requiring emergency surgery is complicated by the fact that these agents have no specific antidote. In this setting, surgery should be deferred for at least 12 hours if at all possible; given NOACs’ short half-lives, this should allow for some mitigation of bleeding risk.³⁰ If delaying surgery is not possible, expert opinion suggests the use of oral activated charcoal or hemodialysis; prophylactic administration of fresh frozen plasma (FFP) or prothrombin complex concentrates (PCCs, discussed below) is not recommended in the absence of major bleeding.³⁰

HERBAL SUPPLEMENTS AND COMPLEMENTARY MEDICINE

The use of herbal supplements and complementary medicine has seemingly exploded in popularity recently and warrants mentioning since many of these naturopathic treatments can have profound hematologic effects. Herbs, such as thyme and rosemary have been shown to have a direct inhibitory effect on platelets.³¹ Fish oil, an omega-3 polyunsaturated fatty acid, may affect platelet aggregation and/or vitamin K-dependent coagulation factors. Omega-3 fatty acids may lower thromboxane A₂ within platelets as well as decrease factor VII levels,³² while garlic, ginger, and Gingko biloba have all been associated with platelet-dysfunction-induced bleeding.³³ Given the many possible antiplatelet and anticoagulant effects of such alternative medicines, not to mention their myriad interactions (both known and unknown) with other drugs, it is therefore prudent to inquire about any supplements patients may be taking prior to cardiac surgery. It is our practice to have patients stop all such remedies 7 days prior to surgery.

Increasing Preoperative Red Blood Cell Mass

Increasing RBC mass prior to surgery is another component of blood conservation, as decreased preoperative hemoglobin or hematocrit levels have been shown to be powerful predictors of the need for transfusion as well as significant risk factors for early and late mortality.^25,34 Such optimization necessitates diagnosing and treating preoperative anemia. Since iron-deficiency anemia is common in the cardiac surgery population, iron supplementation can restore hemoglobin concentrations to adequate levels (eg, 13 g/dL or above), thereby decreasing transfusion risk.

The use of recombinant human erythropoietin (EPO) has been studied as an additional means of improving RBC mass in anemic patients preparing to undergo cardiac surgery. A meta-analysis examining such preoperative EPO administration demonstrated that its use prior to cardiac surgery was associated with a significant reduction in the risk of exposure to allogeneic blood.³⁵ While there are some concerns that chronic EPO use may carry an increased risk of thrombotic complications, a recent randomized, blinded trial of preoperative high-dose EPO of short duration showed a 56% decrease in the relative risk of exposure to blood products in patients undergoing off-pump CABG following preoperative EPO versus those not given EPO; no adverse effects were observed.³⁶ The contemporary STS/SCA guidelines state that it is reasonable to use preoperative EPO and iron several days prior to cardiac surgery in patients with anemia, in those who refuse transfusion (eg, Jehovah’s Witness), or in patients who are at high risk for postoperative anemia.²⁶

Preoperative Autologous Blood Donation

Preoperative autologous blood donation (PABD) remains an option at selected centers for minimizing patient exposures to allogeneic blood. Although this technique has been in practice since the 1960s, its use in cardiac surgery did not achieve widespread acceptance until the 1980s, when the rise of HIV led to increased interest in PABD as a way of reducing allogeneic transfusions. Unfortunately, the acuity of most cardiac operations precludes the routine use PABD as there must be sufficient time between donation and surgery to allow for the regeneration of the patient’s RBC mass. In general, this time is a minimum of 2 weeks per unit of blood donated. What is more, patients must have enough reserve to tolerate the ensuing transient anemia, further limiting PABD’s utility. PABD is only an option for the elective, stable preoperative patient; additionally, there are numerous relative contraindications to PABD, including the presence of left main disease, critical aortic stenosis, congestive heart failure, severe coronary artery disease with ongoing ischemia, active endocarditis, and baseline anemia.

Most of the randomized studies evaluating PABD involve small sample sizes and level A evidence of its benefits are lacking³⁷; nevertheless, a recent case-control study examining the technique showed a lower incidence of allogeneic transfusions among those patients who donated preoperatively compared to those who did not (but 20% of the donated blood products were discarded postoperatively).³⁸ As such, the most recent STS/STA guidelines do not directly endorse PABD; they mention PABD only in the context of EPO use, wherein the guidelines state EPO additionally may be considered to restore RBC volume in patients undergoing PABD.²⁶ For reasons of practicality and cost effectiveness, PABD has largely been supplanted by numerous perioperative blood conservation strategies, which we discuss in the next section.

Cardiopulmonary Bypass Considerations

Numerous CPB techniques have been validated to reduce blood loss and to protect against transfusions during and after cardiac surgery. These strategies, which we discuss in turn, require coordination between the operating surgeon and perfusionist and constitute another vital component of any comprehensive blood conservation program.

ACUTE NORMOVOLEMIC HEMODILUTION

Acute normovolemic hemodilution (ANH) involves the removal of 1 to 2 units of whole blood (target hematocrit of 25–30%) from the patient immediately before, or during, surgery (but prior to CPB initiation) while simultaneously replacing it with crystalloid or colloid to maintain normovolemia. The theoretical basis for ANH (also known referred to as intraoperative autologous donation) is that this lowering of the hematocrit results in fewer RBCs being lost when the patient subsequently bleeds during the course of the operation. Moreover, the removed blood is spared the effects of hemodilution from CPB and is shielded from the inflammatory response of blood cells to the bypass circuit. The collected blood—rich in valuable components (eg, platelets, coagulation factors)—is then infused back to the patient after separation from CPB and heparin reversal.

The amount of blood that an individual patient is capable of donating via ANH depends strictly on the patient’s own physiologic parameters, estimated blood volume (based on height-weight nomograms), and hematocrit. Figure 14-1 is our nomogram for allowable ANH blood drainage—this conservatively estimates the volume of blood that can be removed to achieve a hematocrit of 24% or greater (based on a 1000 mL CPB prime volume).

ANH is relatively contraindicated in patients with preoperative anemia, unstable angina, and those with ejection fractions less than 30%.³⁷ Additionally, evidence supporting its use is conflicting: a few prospective studies showed a significant decrease in allogeneic blood product use,^39,40 others showed no benefit from ANH,^41,42 while a meta-analysis and review found only a modest benefit.⁴³ Contemporary guidelines state that ANH may be considered as part of a multipronged approach to blood conservation in selected (ie, nonanemic) patients, but note that its usefulness is not well-established.^26,44

RETROGRADE AUTOLOGOUS PRIMING

Low hematocrits during CPB were shown to have detrimental effects on end-organ function and cognitive outcomes.^45-49 In retrograde autologous priming (RAP), the CPB circuit is primed with the patient’s own whole blood, thereby minimizing circuit-induced hemodilution (which results from crystalloid priming) and ideally reducing the subsequent need for allogeneic transfusions. Specifically, blood from the aorta is allowed to flow retrogradely through the arterial arm of the bypass circuit, displacing portions of the crystalloid prime. Once the desired prime volume is displaced, a similar procedure can be applied to the venous line (so-called venous antegrade priming) to remove additional crystalloid from the circuit.

RAP was shown to reduce hemodilution and allogeneic transfusions (as compared to conventional CPB priming) in a large, retrospective study⁵⁰ as well as in a recent meta-analysis⁵¹; however, the latter showed that RAP had no effect on clinical outcomes like ventilator hours or length of stay.⁵¹ RAP is currently endorsed by several societal guidelines as a method of reducing allogeneic blood transfusions during on-pump cardiac surgery.^26,44