Danny M. Takanishi Jr.
It is of critical importance that clinicians who provide care to surgical patients understand the dynamic interplay of the coagulation and fibrinolytic systems. An appreciation for the relevant underlying biologic mechanisms is central to the diagnosis and appropriate management of patients who present with bleeding diatheses in both the operative and the post-operative setting. Surgery provides the most significant challenge to the integrity of the hemostatic system, and the fidelity of the coagulation system serves as the homeostatic defense mechanism that abrogates the proclivity for bleeding in this context.
Initial Appraisal of Hemostatic Impairment: General Overview and Immediate Concerns
To expeditiously manage a critically ill patient experiencing life-threatening hemorrhage, it is crucial that a defined approach for prompt recognition of the underlying cause is used. First, attention is directed toward stabilizing the patient. This includes securing an adequate airway, ventilation, vascular access, and restoring intravascular volume. Second, there should be immediate dialogue with individuals involved in the intraoperative management of the patient. This may provide pertinent information regarding the intraoperative course of the patient, specifically regarding any observed characteristics of bleeding that may suggest an underlying coagulation disorder. A generalized slow oozing of blood from raw surfaces (often termed “nonsurgical bleeding”) is often a manifestation of a systemic disorder of hemostasis (1,2,3). On the other hand, bleeding related to technical factors that can be associated with the conduct of any surgical procedure (often termed “surgical bleeding”) is localized and is the most common cause of post-operative bleeding. Dialogue with the surgical team should include consideration for re-exploration. The evolving coagulopathy may proceed swiftly in critically ill patients. Qualitative abnormalities of platelet function, depletion of both platelets and plasmatic coagulation factors, and hypothermia are major contributors to the underlying pathophysiology responsible for the coagulopathy that manifests in the critical care setting (2,3,4,5,6,7,8,9). Moreover, previously undiagnosed, rare, at times clinically latent, congenital coagulation disorders may be unmasked by the physiologic stress resulting from a surgical procedure. Familial coagulation defects that may be encountered in a bleeding postoperative patient include von Willebrand disease, factor VIII deficiency (hemophilia A, or classic hemophilia), and factor IX deficiency (hemophilia B or Christmas disease) (10,11,12,13,14).
The initial phase of assessment must include an immediate reconciliation of intravascular volume status to appropriately guide resuscitation and to assist in decision making for early operative intervention. Although signs of an expanding hematoma or saturated dressings are indicative of localized bleeding, there is considerable overlap, and coagulation system defects may present in a similar fashion postoperatively. In terms of management, however, early recognition of the clinical symptoms and signs attributable to hypovolemia (restlessness, anxiety, shortness of breath, pallor, tachycardia, and oliguria) is compulsory. Adjunctive estimates of intravascular volume status may be obtained using both noninvasive and invasive monitoring, as is common in the critical care setting. Systemic hypotension is a late sign of significant hypovolemia, and expeditious resuscitation is paramount to avert the disastrous consequences of an unrecognized and inappropriately triaged patient with ongoing, potentially life-threatening hemorrhage.
It is important to differentiate the inherited (primary) coagulation disorders, which are associated with a history of bleeding diatheses, from the more commonly acquired (secondary) coagulation disorders, which are the consequence of pathologic conditions and numerous medications (1,3,15). This nomenclature (primary vs. secondary) is distinct from and should not be confused with the traditional nomenclature used to describe the hemostatic process itself. In the formation of a stable clot, the hemostatic process was classically described as comprising two phases, a primary phase (also called primary hemostasis) and a secondary phase (also called secondary hemostasis). The primary phase of hemostasis involves vascular or tissue injury, initiating platelet adhesion and aggregation to form the platelet plug. The secondary phase involves the activation of the plasmatic coagulation protein cascade (both the extrinsic and the intrinsic systems), which results in formation of the stable fibrin clot. Disorders involving platelet number or function, or vascular interactions, are classified as disorders of primary hemostasis, and disorders involving the plasmatic coagulation factors are classified as disorders of secondary hemostasis. This serves the purpose of an operational definition, since in vivo these events are not separate processes but highly integrated.
Hemostatic disorders, whether primary (inherited) or secondary (acquired), can both be manifest by diffuse bleeding from the operative site, puncture wounds, vascular access sites, or traumatized tissue outside of the operative field. Surgical and postsurgical bleeding may therefore result from either quantitative (thrombocytopenia) or qualitative (abnormal function) platelet disorders.
P.1130
Thrombocytopenia may be further attributed to a decrease in production of platelets (aplastic anemia, hypoplastic bone marrow, chemotherapy, space-occupying lesions of the bone marrow as seen with malignancy), ineffective production (vitamin B12 or folic acid deficiency states), platelet sequestration (primary or secondary hypersplenism), increased destruction or consumption of platelets (hemorrhagic conditions; microangiopathic processes such as thrombotic thrombocytopenic purpura [TTP]; disseminated intravascular coagulation [DIC]; hemolytic-uremic syndrome [HUS]; immune destruction due to antiplatelet antibodies, such as in posttransfusion purpura or idiopathic thrombocytopenic purpura), or dilution of circulating platelet volume (massive blood transfusion) (1,2,3,4,6,7,8,9,16,17,18).
Qualitative platelet disorders may be inherited (Bernard-Soulier syndrome, abnormal release mechanism, Glanzmann thrombasthenia, storage pool disease, or von Willebrand disease) or acquired (uremia, drug interactions, myeloproliferative disorders). Disorders involving plasmatic coagulation factors are generally caused by either a decrease in production of clotting factors (commonly associated with hepatic insufficiency, cirrhosis, vitamin K deficiency, obstructive jaundice) or by an increase in consumption of circulating coagulation factors (such as in DIC) (3,6,7,8,9,18,19). Coagulopathy manifested in the critically ill patient is often a result of a combination of both platelet and plasmatic coagulation factor defects. Well-known examples include obstructive biliary tract disease or chronic liver disease, which results in diminished production of coagulation factors given that the liver is the major site of synthesis of all coagulation factors with the exception of factor VIII. In addition, platelet sequestration may occur from secondary hypersplenism (18). Massive trauma often results in decreases of both platelets and coagulation factors as a result of consumption secondary to ongoing bleeding or hemorrhage (4). If shock and acidosis develop, there is further decrease in coagulation factor synthesis due to impairment in liver function that results from low perfusion, in addition to impairment of both coagulation factor and platelet function from the acidemic state (20,21). Additionally, well-described sequelae of massive blood transfusion, often associated with multiple trauma and hemorrhagic shock, is the development of a coagulopathy (22). This results from a combination of dilution of plasmatic coagulation factors and platelets, chelating of ionized calcium (a necessary component of the coagulation cascade) by the anticoagulant (citrate) present in banked blood, acidosis that often attends the underlying disease, hypothermia, or the condition that made massive blood transfusion necessary in the first place. Hypothermia significantly impairs both platelet and plasmatic coagulation function (5,21). Last, transient platelet dysfunction, responsive to desmopressin (DDAVP, 1-desamino-8-d-arginine vasopressin), is a well-recognized phenomenon in patients after cardiopulmonary bypass (23,24,25,26).
In parallel with the ongoing resuscitation of the bleeding patient, attention must be directed toward expeditiously determining the need for surgical re-exploration. All bleeding patients must be considered candidates for reoperation. A comprehensive assessment must take into account the degree and the duration of active bleeding, the anatomic site of involvement, and the potential for additional morbidity or mortality (for example, evolving acidosis, myocardial ischemia, diminished mental status, oliguria, or a progressive neck wound hematoma with impending acute airway obstruction). Physiologic reserve is also a compelling variable to consider, for patients with significant comorbidities, such as the elderly, pediatric, obese and diabetic patients are attended by less reserve. Hence, vigilant observation to recognize early postoperative bleeding is crucial, before hemodynamic instability and shock become manifest. In these patients, consideration for early operative intervention may be necessary because this cohort may not be able to readily tolerate even mild degrees of anemia, hypovolemia, and hypoperfusion. Consideration of other modalities of hemorrhage control (interventional radiology and embolization) may be entertained in appropriate situations.
An important caveat is that not every instance of bleeding requires surgical intervention, and conversely, patients who may initially appear to have clinically self-limited bleeding postoperatively may still require reoperation for control of bleeding. Major pelvic fractures are contemporaneously managed by either fracture stabilization or arteriography with embolization of the culprit bleeding pelvic vessels. In contrast, the natural history of a contained perioperative vascular anastomotic dehiscence is characterized by the evolution of a false aneurysm. In this setting, immediate re-exploration with operative repair is indicated to avert the potentially catastrophic consequence of free rupture and death from exsanguinating hemorrhage.
Evaluation of Bleeding: Clinical History and the Utility of Diagnostic Testing
A detailed history and physical examination is the most important preliminary step in elucidating the cause of surgical bleeding and should be done simultaneously with resuscitative efforts (1). Collateral history from family members and previous medical records is a helpful adjunct to determine if a primary, congenital defect in the coagulation system is present. A history of easy bruisability, excessive gingival bleeding after brushing of teeth, bleeding diathesis with dental extractions, hypermenorrhagia, frequent spontaneous epistaxis, melenic stools or spontaneous hematuria, petechiae or purpura, hemarthroses, and a family history of bleeding disorders may indicate a congenital or familial coagulation disorder, such as hemophilia A or B or von Willebrand disease. The family pedigree may provide important clues as to the presumptive disease process based on the pattern of inheritance, whether autosomal or sex linked, dominant or recessive. Suspicion for a congenital disorder of coagulation is further raised by a history of blood transfusions required for common ambulatory procedures such as dental extractions, circumcisions, tonsillectomies, or biopsies. Due to differences in gene penetrance, not all individuals afflicted with inherited coagulation disorders are diagnosed at an early age, and clinically latent, attenuated bleeding disorders may be unmasked when confronted by a major surgical procedure or trauma. A past history of liver disease or heavy ethanol consumption should alert the clinician to the possibility of acquired plasmatic factor deficiencies, in addition to thrombocytopenia resulting from secondary hypersplenism with platelet sequestration.
A thorough medication history is essential (including soliciting information on the use of dietary supplements or herbal tonics, and over-the-counter medications), to determine if the patient has been on any medication that interferes with hemostasis (common examples include aspirin, other nonsteroidal anti-inflammatory agents, ticlopidine, clopidogrel, or semisynthetic penicillins) (16,23,24,25,26). Popular supplements that may aggravate bleeding are ginkgo, garlic, ginger, ginseng, feverfew, and vitamin E. A history of anticoagulation therapy is equally important in this regard (19,27). Nutritional assessment is paramount, and careful evaluation for the presence of a vitamin K deficiency is obligatory, as this may occur in patients on parenteral nutrition or with cancer cachexia in those with malignancies. Other variables that may affect the integrity of the coagulation system include previous irradiation, renal failure, and sepsis, which affect the coagulation cascade at multiple points.
Physical examination often provides an index of the severity and the extent of the disease and may provide additional clues that assist in distinguishing localized surgical bleeding from systemic bleeding resulting from a coagulopathy. For example, the presence of petechiae, purpura, and mucosal bleeding is often indicative of thrombocytopenia, a qualitative functional disorder of platelets, or increased vascular fragility. Ecchymoses or spontaneous, nontraumatic hemarthrosis is consistent with plasmatic coagulation factor abnormalities or deficiencies such as hemophilia. Both platelet and plasmatic coagulation disorders are associated with hematomas. Hepatic insufficiency or failure can be presumptively identified by recognizing jaundice, ascites, angiomas, palmar erythema, asterixis, congestive splenomegaly, and testicular atrophy. Splenomegaly itself may also be associated with hematologic dyscrasias and malignancies associated with hemostatic abnormalities (lymphomas and leukemias). Connective tissue or collagen vascular disorders that result in increased vascular fragility may manifest with petechiae, joint abnormalities, and a history of delayed wound healing. These conditions focus attention on the increased risk for a perioperative bleeding complication, which underscores the need for vigilance both intraoperatively and postoperatively. Last, the possibility of sepsis as an underlying cause for the development of a coagulopathy must always be entertained in the postoperative, critically ill patient (9).
The establishment of a definitive diagnosis of a coagulation disorder rests on selective use of a limited battery of laboratory tests guided by information derived from the history and physical examination. These assays are selected to broadly screen the hemostatic system. These tests are, for the most part, automated (except for the bleeding time), readily available, and amenable to point-of-care testing methodology. The tests most commonly used include the template bleeding time, quantitative platelet count, prothrombin time (PT), activated partial thromboplastin time (PTT), fibrinogen level, and thrombin time (TT). The template bleeding time is the only test that screens for qualitative platelet function abnormalities, a frequent cause of abnormal bleeding. This test also evaluates platelet number and vascular fragility, demonstrating abnormal prolongation in thrombocytopenic states and in conditions associated with increased vascular fragility (examples include connective tissue disorders such as senile purpura, Ehlers-Danlos syndrome, steroid-induced purpura, or Marfan syndrome; scurvy; amyloidosis; or hereditary hemorrhagic telangiectasia/Osler-Weber-Rendu disease). Therefore, if both the platelet count and the template bleeding time are normal, the presumptive differential diagnosis is directed toward a plasmatic coagulation factor abnormality. Measurement of both the PT and PTT will serve to further define the abnormality, given that each test is more sensitive to changes in procoagulants in the initial phases of the extrinsic and intrinsic pathways, respectively. The PTT provides a global measure of the activity of factors XII, XI, IX, and VIII in addition to the common pathway factors shared by the extrinsic system (factors X, V, II, and I) and, therefore, identifies many of the inherited disorders of bleeding, typically a deficiency of factor VIII (hemophilia A), IX (hemophilia B), or von Willebrand factor, vWF (von Willebrand disease). It is worth noting that factor XII deficiency is not associated with any significant bleeding tendency despite abnormal prolongation of the PTT. Enzymatically active vWF is a necessary cofactor for optimum functioning of the factor VIII procoagulant protein. The PT exclusively evaluates factor VII, which is one of the vitamin K–dependent factors (in addition to factors II, IX, and X). Therefore, the PT is prolonged in individuals on Coumadin (warfarin) therapy, and this test is used to measure therapeutic efficacy of this form of oral anticoagulation therapy. Liver disease is another common cause for a prolonged PT, perhaps most notably because factor VII has the shortest half-life of the plasmatic coagulation factors. Liver disease, by virtue of the diminished synthesis of all plasmatic coagulation factors (with the exception of vWF) also results in prolongation of the PTT in addition to the PT. Due to the sensitivity of factors XII, XI, IX, VIII, and X to the effects of heparin (in the presence of antithrombin III), heparin therapy primarily is reflected by abnormal prolongation of the PTT. In conditions associated with prolongation of both the PT and the PTT, quantitative measurement of the fibrinogen (factor I) level may be useful to determine if the coagulation disorder is a result of a defect or deficiency of multiple coagulation factors, which may be associated with liver disease, sepsis, and DIC, or if an inhibitor is present (3,6,7,8,9,18). Examples of circulating inhibitors include heparin, paraproteinemias associated with monoclonal gammopathies (multiple myeloma or Waldenstrom macroglobulinemia), fibrin/fibrinogen degradation products, or other circulating anticoagulants (such as antibodies to factor VIII or IX, and the lupus anticoagulant). Factor levels reflect a dynamic equilibrium, and fibrinogen is an acute-phase protein that increases in response to stress. Fibrinogen consumption may not be apparent based solely on measurement of levels because of the propensity for the liver to increase synthesis of this protein in response to the same stressors responsible for fibrinogen consumption. This fact is important to bear in mind for proper interpretation of laboratory results. The thrombin time is a qualitative measure of fibrinogen levels and will be prolonged if the fibrinogen level is less than 100 mg/dL, by the presence of dysfibrinogenemia associated with liver disease, and also by circulating inhibitors similar to the PT and PTT (28). Its most useful role clinically is in the detection of circulating heparin not detectable by changes in the PTT.
The therapeutic approach to hemostatic abnormalities must be reasonably guided by the patient's clinical condition and the outcome of appropriately selected laboratory tests, as outlined above (14). An important concept is not to fall behind in factor replacement. Due to the lag time in receiving laboratory values, in clinical conditions where the patient is actively bleeding, initiation of fresh frozen plasma (FFP) and platelets should be started early with red cell transfusion. In patients with known warfarin use and a small bleed in an uncompromising area (skull), FFP should be used before patients develop increasing intracranial hemorrhage. In general, FFP is the most commonly used blood component for plasmatic coagulation abnormalities since it provides all necessary coagulation factors in concentrations that approach those found in normal plasma (3,10,11,14). Cryoprecipitate, a component made from thawing FFP in the cold under specialized conditions, is rich in factors XIII, VIII, I, vWF, and fibronectin. It has been used in factor VIII–deficient states (hemophilia A) if virally inactivated plasma-derived or recombinant factor VIII concentrates are not readily available; in von Willebrand disease recalcitrant to desmopressin and if virally inactivated plasma derived factor VIII concentrate rich in vWF is not readily available; in hypofibrinogenemic or dysfibrinogenemic conditions; in hemorrhage associated with massive blood transfusions; in DIC; in factor XIII deficiency (FFP is also used); in bleeding uremic patients with qualitative platelet function defects unresponsive to desmopressin therapy; or as a topically applied surgical sealant (“fibrin glue”) (10,14,29). For specific coagulation disorders, such as hemophilia A, hemophilia B (deficiencies of factor VIII and IX, respectively), and von Willebrand disease, virally inactivated plasma-derived factor concentrates are commercially available, as are recombinant products, and these are the blood components of choice if available for use. These concentrates (and the recombinant product) also have a lower risk for virally transmitted disease because of a processing procedure that results in viral inactivation. For deficiencies of the vitamin K–dependent factors (II, VII, IX, X), prothrombin complex concentrates are also commercially available, as is recombinant factor VII for isolated factor VII deficiency states. Platelet concentrates are used for both quantitative (thrombocytopenias) and qualitative platelet disorders associated with significant bleeding (3,10,12,14). Patients transfused with any blood component, whether FFP, cryoprecipitate, plasma-derived factor concentrates, or platelet concentrates, incur similar risks of complications attributed to all types of blood component therapy such as febrile and allergic reactions, transmission of blood-borne pathogens (human immunodeficiency virus, hepatitis B and C), immunosuppression, transfusion-related pulmonary injury, and hemolytic transfusion reactions if a substantial volume of ABO-incompatible units are transfused. Therefore, judicious use of blood component therapy must always be adhered to.
Causes of Post-Operative Bleeding
Post-operative bleeding, often termed local hemostatic failure or surgical bleeding, is a known potential complication of any surgical procedure. When associated with evolving hypovolemia, mental status changes, restlessness, anxiety, tachycardia, dyspnea, and oliguria are commonly associated manifestations. Hypotension is a late finding, and aggressive attempts must be made to avert this serious consequence with expeditious concurrent resuscitation as identification of the source of bleeding is confirmed. The vast majority of patients affected typically present within the immediate perioperative period. Subtle signs may be evident in the postanesthesia care unit, and generally become apparent within the first 8 hours after surgery. A high index of suspicion is necessary to render an early diagnosis of post-operative bleeding, and meticulous attention to look for any evidence of bleeding must be applied, given that many signs of evolving hypovolemia are nonspecific and may also be observed in nonbleeding patients after major thoracic or abdominal procedures (tachycardia associated with post-operative pain, anxiety and restlessness, mental status changes secondary to narcotic analgesic administration, or oliguria resulting from anticipated third-space fluid sequestration after major abdominal surgery). Blood in the peritoneal cavity ordinarily does not result in a significant inflammatory response unless associated with secondary bacterial contamination and therefore is not associated with obvious peritoneal signs. On occasion, localized symptoms may be elicited that are attributable to irritation caused by a collection of blood, exemplified by the Kehr sign. This is referred pain to the right shoulder ascribed to an accumulation of blood under the right hemidiaphragm. Serial hemoglobin and hematocrit levels may assist in determining the degree of bleeding, but isolated, single values can be difficult to interpret. It is difficult to quantitatively account for the effects of isotonic fluid sequestration (third-spacing) after major abdominal procedures that may result in hemoconcentration and elevated hemoglobin and hematocrit levels, or the effects of isotonic fluid administration in the perioperative period that may contribute to hemodilution and lower hemoglobin and hematocrit levels.
Certain surgical procedures are not associated with exsanguinating hemorrhage or hypotension but can be life threatening. Prototypic examples include neck operations for endocrine diseases (thyroid, parathyroid surgery), lymphadenectomies (radical neck dissections), major composite resections for tumors of the neck, and carotid surgery. From a pathophysiologic perspective, an expanding neck hematoma results in airway obstruction from both mechanical compression and from mural edema caused by lymphatic obstruction. The cause is frequently venous bleeding, compounded by hypertension and liberal preoperative use of antiplatelet medications in those undergoing carotid surgery. Acutely reopening the incision and evacuating the hematoma is often life-saving, although endotracheal intubation or cricothyroidotomy may be required as a temporizing measure until the airway edema resolves. Nearly all patients who develop neck hematomas require operative re-exploration. It is common practice to place drains in the operative field at the time of surgery prior to closure, and surgical procedural anthologies are replete with instructions substantiating this approach. Caution must be exercised in interpreting drain output as an accurate index of early perioperative bleeding, and it is correspondingly crucial to recognize that placement of a drain is not an appropriate substitute for ensuring meticulous surgical hemostasis (30).
There is no single criterion available to direct re-exploration for control of post-operative bleeding, and this decision is based on considering a number of variables. The timing of the active bleed relative to the operative procedure, its duration, its rate, the potential for additional morbidity, the patient's age as a surrogate for physiologic reserve, and other comorbid diseases (such as underlying cardiac or pulmonary disease, renal disease, diabetes, or obesity) must all be taken into consideration when deciding on the need for reoperation. Timing of re-exploration is of significant concern in conditions associated with limited or poor physiologic reserve as these patients are often quite ill and require judicious resuscitation and expeditious definitive surgical intervention prior to the inception of irreversible shock. The most conservative treatment is to return to the operating room with early control of surgical bleeding. Often times the need to return emergently to the operating room is quite obvious, as in the case of exsanguinating hemorrhage resulting from a coronary artery bypass graft dehiscence attended by brisk bleeding from the mediastinal drain. There are two caveats for consideration. First, it is desirable to correct any coagulopathy prior to returning to the operating room, and in some instances with minimal or mild bleeding this may be all that is necessary. It is important to recognize, however, that situations characterized by exsanguinating hemorrhage from failure of local surgical hemostasis may not allow for correction of the coagulopathy because of rapid ongoing consumption of coagulation factors and platelets. In these instances, operative intervention is paramount, and the decision to reoperate must not be unduly delayed while awaiting normalization of coagulation parameters. Second, as stated earlier, some subscribe to the notion that drains placed at the time of surgery are a useful adjunct to alert the surgical team to early signs of postoperative bleeding and to gauge the amount and rate of bleeding when it does occur. Caution with this practice must be promulgated, as it is a well-accepted observation that the absence of blood in a drain is not conclusive evidence that bleeding is not occurring, because the drain tip may be dislodged or may have migrated from its original position and this may not be readily apparent externally, or the drain may be obstructed with clot.
There are certain circumstances where operative re-exploration is obligatory, despite the fact that the bleeding may be self-limiting. A vascular anastomosis with a contained leak, even if seemingly small and hemodynamically inconsequential, must be repaired to avert the consequences of false aneurysm formation and later rupture. Small contained leaks may be difficult to recognize and pose a diagnostic dilemma. A duplex study may be helpful in further elucidating anastomotic integrity, but an arteriogram should be performed if any doubt exists.
Last, the surgeon must also be cognizant of the fact that there are some instances where bleeding is optimally addressed without surgical intervention. The prototypic illustration is the severe pelvic fracture with signs of ongoing hemorrhage, as alluded to earlier. In accordance with the Advanced Trauma Life Support protocol, associated injuries and additional sources of obvious bleeding must first be excluded. Once hemoperitoneum and hemothorax have been excluded, the pelvis is stabilized by external fixation, followed by arteriography with embolization of any bleeding pelvic or retroperitoneal blood vessels.
Disorders of Plasmatic Coagulation Factors: Overview
Hemostasis involves a complex interplay between elements of the vascular endothelium, plasmatic coagulation factors, platelets, and the fibrinolytic system. Primary isolated disorders of the fibrinolytic system as a cause of major bleeding are rare in the critically ill patient and are covered elsewhere (31,32). Disorders of the clotting system can be broadly classified as congenital or acquired, and a comprehensive history and physical examination often assists in determining the nature of a coagulation disorder in this regard. Acquired coagulopathies are relatively common in the post-operative ICU patient, and these disorders can be operationally separated in those involving platelets, the plasmatic coagulation factors, or both (3). Thus, post-operative bleeding can be a result of quantitative abnormalities (thrombocytopenia), due to decreased production, ineffective thrombopoiesis (production), sequestration, increased destruction or consumption, and dilution; or qualitative platelet function abnormalities. Post-operative bleeding can also occur with acquired disorders of the coagulation system, which encompasses decreased production, impairment of function, or increased consumption of coagulation factors.
Quantitative Platelet Disorders: Thrombocytopenia
Thrombocytopenia, defined as a platelet count less than 140 × 109/L, results from decreased production, ineffective thrombopoiesis, sequestration particularly in the spleen, increased destruction or consumption, and dilution of circulating platelets associated with massive blood transfusion (1,10,12,14). Petechiae, purpura, and mucosal oozing is characteristic of thrombocytopenic states, although these findings may also be observed in qualitative platelet disorders and in conditions associated with increased vascular fragility. Platelet counts of 50 × 109/L or higher are generally considered adequate for surgical hemostasis in the absence of an associated qualitative functional defect, but below 20 × 109/L there is increased risk for spontaneous hemorrhages (10,12,14). Particularly lethal in this regard are those involving the central nervous system.
Cytotoxic chemotherapy and radiation therapy (total body) produce thrombocytopenia by suppression of bone marrow megakaryocytes, the progenitor cell for platelets. Together these are the most common causes of bleeding in patients undergoing therapy for malignancies. Marrow aplasia, hypoplasia, and space-occupying diseases of the bone marrow (such as metastatic carcinoma, leukemias, lymphomas) also result in decreased production of platelets. Certain drugs, including alcohol, have been associated with decreased production of platelets via a direct toxic effect on megakaryocytes. Ineffective thrombopoiesis is a characteristic trait of megaloblastic anemia resulting from either vitamin B12 or folate deficiency. Although there is an increase in the megakaryocytic mass, platelet production is impaired. Notwithstanding, hemorrhagic diatheses manifest in a few of these individuals (33).
The most frequent cause of thrombocytopenia resulting from increased destruction of circulating platelets is post-operative infection (9,34). Significantly, thrombocytopenia may be the first presenting sign of an occult infection and may herald impending sepsis. Consequently, in the post-operative setting, thrombocytopenia of unclear cause must promptly direct attention toward uncovering a potential source of occult sepsis.
Thrombocytopenia caused by increased destruction of platelets is also observed in microangiopathic hemolytic states, such as in thrombotic thrombocytopenic purpura (TTP) and hemolytic-uremic syndrome (HUS) (35,36). Mechanical injury occurs when platelets traverse the small capillary beds in the peripheral circulation. TTP is characterized by fever, fluctuating neurologic symptoms (headaches, confusion, seizures, or coma), and acute renal failure, in addition to a microangiopathic hemolytic anemia and thrombocytopenia. HUS manifests similarly to TTP, with the notable exceptions that the pediatric population is more commonly affected, neurologic manifestations are minimal, but the renal impairment is more pronounced. Central to the treatment of both these entities is supportive care, with particular attention given to management of the renal dysfunction (35,36,37). Plasma exchange is often efficacious in treating these diseases, and hemodialysis may also be required in some instances for support of renal failure.
Immune-mediated destruction of platelets is observed in several clinical conditions. Alloimmune antibodies are believed to account for posttransfusion purpura observed primarily in women, who may have been previously immunized by fetal-derived platelets since there is a significant association with a prior history of pregnancy (38). Immunizations to a number of candidate alloantigens have been reported in the literature, the most common being the PLA1 antigen. The antibodies induced are generally of the IgG class and therefore are also able to cross the placenta, as a described cause of neonatal thrombocytopenia. The purpura becomes apparent approximately 7 to 10 days after blood transfusion, presumably attributed to an anamnestic response, and can last several months. The population at risk has been estimated to be approximately 1% to 3%, and the condition tends to be self-limiting and responds to intravenous immune globulin. Idiopathic thrombocytopenic purpura is one of the more common examples of immune-mediated platelet destruction (17). It tends to occur in otherwise healthy individuals, and both an acute and a chronic form have been described. Mechanistically, platelets coated with autoantibodies are removed by the reticuloendothelial cells in the spleen (and to an extent in the liver). The diagnosis is one of exclusion after other causes of thrombocytopenia have been ruled out. A similar mechanism may account for the thrombocytopenia associated with collagen vascular diseases, such as systemic lupus erythematosus, lymphoreticular diseases, and in some infectious diseases, such as infectious mononucleosis or human immunodeficiency virus infections. It is noteworthy in this regard that the acute form, often observed in the pediatric population, is often preceded by a viral syndrome. Splenectomy is required in a third of patients if immune globulin, corticosteroids, or plasmapheresis is unable to control the condition (17,39,40). In this condition, significant bleeding may not occur until platelet counts decrease as low as 10 × 109/L because most of the circulating platelet pool consists of younger, more functionally active platelets. The significance of this impacts on the operative approach traditionally adopted during splenectomy. During splenectomy, platelets are hung by the anesthesiologist but not administered until the splenic artery is clamped or splenectomy completed. There is by and large minimal bleeding encountered despite the pronounced degree of thrombocytopenia, and if platelets are infused prior to control of the splenic arterial inflow, the infused platelets will merely be consumed by the spleen and not available for the hemostatic process. Immune-mediated destruction of platelets can also be caused by several drugs that can induce antibodies to platelets via hapten-mediated, or by immune complex–mediated, “innocent bystander” mechanisms. Quinine, amiodarone, sulfa drugs, cimetidine, ranitidine, phenytoin, and semisynthetic penicillins are some examples that may be encountered in the critical care environment (3). Heparin-induced thrombocytopenia is an unusual example of drug-induced thrombocytopenia in this context because a hypercoagulable condition is actually created characterized by thrombotic complications with the manifestation of the “white clot syndrome” (41). This syndrome typically becomes apparent after 1 week of therapy but may present within a few hours after implementing heparin therapy in already sensitized patients. Discontinuation of the offending agent is the appropriate treatment approach central to all causes of drug-induced, immune mediated thrombocytopenia.
Qualitative Platelet Function Disorders
Acquired disorders are the leading cause of qualitative platelet function abnormalities in the critically ill patient. It is vital to be aware that qualitative bleeding disorders are not measured by the standard battery of coagulation tests described above, with the exception of the template bleeding time. This test is not commonly used, given that this does not lend itself to automation and still requires the laboratory technologist to remain at the patient's bedside. Furthermore, this test has been attended by poor reproducibility, particularly in conditions associated with significant peripheral edema. Nevertheless, when a qualitative platelet function abnormality is suspected, the bleeding time is an appropriate first screening test to guide discriminate use of additional testing to further elucidate the underlying cause.
Ingestion of numerous drugs has been associated with inhibition of platelet function (3,16,42). Among these, aspirin is the most well described and best characterized. Aspirin interferes with cyclo-oxygenase–mediated prostaglandin and thromboxane synthesis and has profound effects at multiple steps in the formation of the hemostatic platelet plug. It decreases the platelet response to aggregation in response to collagen, inhibits the second phase of aggregation in response to adenosine diphosphate (ADP) and epinephrine, and irreversibly injures platelets for the duration of their lifespan. It is generally recommended that antiplatelet medications, such as aspirin, ticlopidine, and clopidogrel, are discontinued approximately 7 to 10 days prior to surgery. For nonsteroidal anti-inflammatory agents other than aspirin, ticlopidine, or clopidogrel, some investigators advocate 2 days of abstinence. It is important to recognize that many over-the-counter medications contain aspirin (e.g., Alka-Seltzer, Ecotrin, Anacin) and many patients are not aware of this, so it is imperative to obtain a comprehensive drug history, specifically querying for use of aspirin-containing products. Discontinuing the antiplatelet drug combined with use of desmopressin and platelet transfusions have been beneficial in treating bleeding encountered in these situations (10,12,14,16).
Another common cause of acquired qualitative defects in platelet function is hypothermia (5,21,43). Massive blood transfusions or crystalloid infusions without attention to use of blood warmers, lack of attention to maintaining a warm ambient environment in the operating room, especially during long procedures and for individuals at the extremes of age, and prolonged extrications and exposure time in the field in the patient with multiple traumatic injuries are all too familiar causes of hypothermia. Vigilant awareness must be directed toward maintaining normothermia to avert the potentially disastrous consequences of hypothermia in the already bleeding critically ill patient. Rewarming patients reverses the effects of hypothermia on the hemostatic system.
Renal failure is not uncommon in the critical care setting. In its acute form, bleeding is a common manifestation, most often from the gastrointestinal tract. The underlying mechanism is probably multifactorial, as there clearly is a qualitative platelet function defect related to the degree of uremia, in combination with abnormalities in the plasmatic coagulation system (44,45,46). The presence of acidosis also contributes to both the platelet and coagulation factor dysfunction. The fundamental approach to therapy centers on dialysis, which results in abatement of the bleeding diathesis. Use of desmopressin and cryoprecipitate, as temporizing measures to transiently stop the bleeding while awaiting institution of dialysis, have been reported to be successful. Conjugated estrogens have also been used with some success, albeit the effects are not as rapid but more durable, but any positive outcomes are balanced by undesirable consequences of hormonal side effects. In its chronic form, renal failure is still attended by a mild qualitative platelet function defect, but significant impact on hemostatic homeostasis is usually not seen.
Acquired Disorders of the Coagulation System Decreased Production
Decreased production of circulating plasmatic coagulation factors occurs secondary to liver failure (with the exception of factor VIII), vitamin K deficiency (seen with oral antibiotic usage, which depresses gut flora in the setting of nutritional deficiency; malabsorption syndromes, such as celiac sprue or chronic diarrheal conditions; or obstructive jaundice), and use of warfarin (Coumadin) (18,19). In these situations, therapy is acutely centered on replacement of coagulation factors, most commonly with use of FFP. Vitamin K is administered parenterally in patients with deficient states or to reverse the effects of Coumadin. Appreciable effects on coagulation factor synthesis (in the presence of normal hepatic synthetic function) is not generally seen for 24 to 36 hours after administration of parenteral vitamin K (19,27).
Impaired Function
The effect of hypothermia on antagonizing normal functioning of the hemostatic mechanism globally has been described under qualitative platelet abnormalities (2,5,21,43). It is important to bear in mind that all enzymatic processes in biologic systems are governed to an extent by the necessity to function in an optimal, typically narrow, temperature range. The coagulation factors are enzymes, and therefore function best under normothermic conditions. Platelets, too, function optimally under normothermic conditions. The implication of this is that the clinician must always remain alert to the effects of hypothermia as the origin of a coagulopathy, particularly in the critical care environment. Reliance solely on the values of the PT, PTT, or TT is misleading, understanding that these assays are performed by both manual and automated laboratory methods at 37°C and results may thus fall within the reference range in vitro, despite ongoing coagulopathy clinically. External warming measures include blankets generating heated air, warming of all fluid infusions, heated humidifier in ventilated patients, and warming the environment. In extreme conditions, Gentilello et al. (47) reported that continuous arteriovenous rewarming, which does not require heparinization of the patient and hence does not exacerbate the coagulopathy, improves hemostasis more rapidly than any other method with the exception of cardiopulmonary bypass.
Normal physiologic processes (fibrinolytic system) exist to control for unremitting clot formation and are described in several reviews (1,31,32). Impaired function of the coagulation cascade may be the result of various disorders that are characterized by the genesis of circulating anticoagulants, abnormal protein products, or accumulation of proteinaceous breakdown products that affect the normal function of coagulation proteins. Collagen vascular diseases, such as systemic lupus erythematosus, is one example (48,49). In these patients an antibody is produced (lupus anticoagulant) that affects the coagulation cascade at the juncture of the intrinsic and extrinsic systems, resulting in prolongation of both the PT and PTT in vitro. Paradoxically, these patients tend to be hypercoagulable, and if clinically significant bleeding is noted, it is attributable to associated thrombocytopenia and increased vascular fragility. When there are elevated titers of either the lupus anticoagulant or anticardiolipin antibodies, or both, these patients may present with manifestation of the antiphospholipid antibody syndrome with generalized microvascular thrombosis, thrombocytopenia, gangrene of the extremities, multiorgan failure, and death. Plasmapheresis, anticoagulation, and immunosuppressive therapy serve as the foundation of treatment (48,49). Other commonly acquired inhibitors or circulating anticoagulants include factor VIII inhibitors and factor IX inhibitors, related primarily to prior frequency of transfusion with plasma-derived blood concentrates and alloimmunization (50,51,52). Exogenously administered heparin, the prototype for anticoagulation therapy, binds to circulating antithrombin III and catalyzes its ability to neutralize the action of a number of coagulation factors. The end result is interference with the normal coagulation cascade. Disorders characterized by the production of abnormal globulins, often referred to collectively as the paraproteinemias (associated with multiple myeloma and Waldenstrom macroglobulinemia), also result in interference with coagulation proteins and inhibition of fibrin polymerization. In these conditions the PT, PTT, and TT are prolonged. Fibrinogen/fibrin degradation products also inhibit fibrin polymerization, as does uremia, with prolongation of the PT, PTT, and TT. Treatment of the coagulopathy associated with all of these conditions consists of replacement of deficient coagulation factors when bleeding is dominant and definitive treatment directed at the underlying disease process.
Increased Destruction or Consumption
The most common cause of increased destruction of plasmatic coagulation factors has been variously termed DIC, defibrination syndrome, or consumptive coagulopathy (6,7,8,9). This syndrome is characterized by a hemorrhagic diathesis with unrestrained clotting and fibrinolysis in the vascular microcirculation, initiated by activation of the intrinsic or the extrinsic system, or both (6,7,8,9,37). Release of tissue thromboplastin, from injured tissue or from leukocytes, activates the extrinsic system, whereas damage to vascular endothelium (in addition to releasing tissue thromboplastin) results in activation of the intrinsic system via collagen exposure (8). Exposed collagen initiates platelet aggregation with release of platelet factor III and also activates factor XII directly. The net result is deposition of fibrin in the microvasculature. This results in a microangiopathic hemolytic anemia with fragmentation of red blood cells as they traverse these vascular beds. These fragmented red blood cells, or schistocytes, seen on the peripheral blood smear are a classic finding in this syndrome. Additionally, microthrombi cause stasis and ischemia in a number of capillary beds, manifesting as renal insufficiency or failure with kidney involvement, pulmonary insufficiency with lung involvement, mental status changes with brain involvement, or dermal necrosis with skin involvement. Stasis itself can result in further activation of clotting factors. Fibrin deposition and endothelial wall damage both bring about the release of plasminogen activator, which catalyzes the conversion of circulating plasminogen to plasmin. Plasmin proteolytically hydrolyzes both fibrinogen and fibrin (secondary fibrinolysis), resulting in fibrinogen and fibrin degradation (or “split”) products. These degradation products then interfere with fibrin polymerization through the formation of complexes, further contributing to the hemorrhagic state. Additionally, these degradation products also interfere with platelet function, impairing both adhesion and aggregation. The number of disorders associated with DIC is substantial, but the unifying approach to management is supportive therapy with replacement of coagulation factors and platelets with attention focused on treating the underlying disease process. The end point is normalization of the PT, PTT, TT, and platelet count. Both acute and chronic forms have been identified. In the acute form patients are critically ill, whereas in the chronic form the natural history is more indolent and protracted, and thrombotic complications may be the predominant feature.
In the post-operative patient in the ICU, infection is the principle cause for DIC (9). Several causative organisms have been implicated, including Gram-negative bacteria, such as the Enterobacteriaceae as well as the nonlactose fermenters; Gram-positive bacteria; rickettsial organisms (Rocky Mountain spotted fever); mycotic infections, such as disseminated aspergillosis; parasitic agents, such as malaria; and viruses. The underlying pathophysiology has been best elucidated with Gram-negative infections, with endotoxin (cell wall lipopolysaccharide) triggering the intrinsic system by activation of factor XII directly and by factor XII exposure to subendothelial collagen, as a result of endotoxin-mediated damage to vascular endothelium. Endotoxin may also trigger the coagulation cascade by inducing expression of procoagulant activity in circulating leukocytes, hepatic macrophages, and endothelial cells, and by activating the extrinsic system mediated by the release of tissue thromboplastin from damaged leukocytes and vascular endothelium (53,54,55).
Traumatic injuries (particularly involving brain, bone, or liver), thermal injuries, and severe crush injuries, as well as surgical procedures may produce a consumptive coagulopathy (6,7,8,9,56,57,58). Secondary infection and hemorrhagic shock further serve to aggravate the coagulopathy, especially if acidosis, hypothermia, or tissue ischemia and necrosis develops.
Acute pancreatitis, arising from various causes, may be associated with DIC due to release of enzymes that may directly activate a number of coagulation factors (59). In many instances, there is associated multiorgan dysfunction involving cardiopulmonary, renal, and hepatic function. In addition, pyogenic sequelae, such as the development of infected pancreatic necrosis or abscess formation, may result in DIC attributable to sepsis. Treatment is primarily supportive, with aggressive resuscitation, replacement of deficient coagulation factors if there is associated bleeding, appropriate use of broad-spectrum antibiotics, and surgical debridement and drainage for control of infectious complications.
Obstetric complications can result in some of the most profound and challenging instances of DIC. Well-recognized examples include amniotic fluid embolism, abruptio placentae, retained dead fetus, and eclampsia (6,7,8). In these circumstances, the culprit is massive systemic release of tissue thromboplastin that generates a fulminant course characterized by bilateral renal cortical necrosis to frank cardiopulmonary collapse, shock, multiorgan failure, and, at times, death even if aggressive attempts are made to treat these individuals.
The laboratory diagnosis of DIC is readily established with routinely available tests. The PT, PTT, and TT are all prolonged, and the platelet count is decreased. Depending on the severity of the disease process, fibrinogen may not be detectable. Fibrinogen/fibrin degradation products are elevated, and the peripheral blood smear often reveals the presence of schistocytes. Factors I (fibrinogen), V, VIII, and XIII tend to be markedly depressed. In milder forms, fibrinogen levels may not be significantly decreased, particularly in the presence of adequate hepatic function. Radioimmunoassays of fibrinopeptide A, a by-product of the action of thrombin on fibrinogen, may be useful in these situations to establish the correct diagnosis and management approach (60). Although rare, primary fibrinolysis differs from DIC (where secondary fibrinolysis occurs) in the following ways. In primary fibrinolysis, (a) platelet count is normal, (b) soluble fibrin monomers are not present (measured by the plasma paracoagulation test), (c) schistocytes (red cell fragments) are not seen, and (d) tests for increased levels of plasmin activity are strongly positive (euglobulin clot lysis time, whole blood clot lysis time).
Optimum management of DIC requires aggressive treatment of the underlying disease process and supportive therapy with coagulation factor and platelet replacement. FFP and platelet concentrates are the two most common blood products used as a temporizing measure (10,11,12,14). Stored or banked whole blood is a reasonable source of most clotting factors if the units are less than 24 hours old (10,14,61). The biologic half-life of factors V, VII, VIII, and IX are on the order of 24 hours or less; hence whole blood stored for longer than 24 hours may not provide adequate amounts of these coagulation factors. Heparin had been used in the past in a theoretical attempt to abrogate the clotting cascade, but its contemporaneous use for this purpose is at best controversial, may be contraindicated in the perioperative period, and is not supported by evidence-based data (6,7,8,62). In septic patients with or without DIC, treatment with recombinant activated protein C reduces mortality, and this may be attributable in part to its profibrinolytic, anti-inflammatory, and anticoagulant effects (6,63).
Complex Post-Operative Bleeding Problems
Hepatobiliary Disease
Hepatic parenchymal and biliary obstructive disease results in diverse manifestations of hemostatic abnormalities (18). Extrahepatic biliary obstruction results in diminished absorption of vitamin K due to lack of bile salts necessary for gastrointestinal absorption of lipid soluble vitamins. Decreased synthesis of the vitamin K–dependent factors II, VII, IX, and X occurs with abnormal prolongation of the PT and eventually the PTT.
Parenchymal diseases such as cirrhosis, chronic active hepatitis, fulminant hepatic failure, or metastatic carcinoma impact on the hemostatic system in a heterogeneous manner. Most coagulation factors, naturally occurring anticoagulants (such as antithrombin III), fibrinolysin precursors (plasminogen), and inhibitors of the fibrinolytic system (antiplasmins) are synthesized by the liver. In severe liver disease, acquired dysfibrinogenemia has also been reported (18,28). This impairs polymerization of soluble fibrin monomers and is suggested by a prolonged TT on purified fibrinogen, which is generally done in a research laboratory. The liver also removes activated coagulation factors from the circulation, but it is speculative to conclude that this results in a coagulopathy by itself, despite the fact that this increased consumption of activated factors by the liver lowers coagulation factors already depressed by decreased production in a diseased liver. Clearance of fibrinogen/fibrin degradation products is reduced in chronic liver disease. These breakdown products inhibit both fibrin polymerization and platelet function and thus contribute to a defective hemostatic system, as discussed earlier. Thrombocytopenia occurs secondary to hypersplenism, potentially exacerbated by vitamin deficiencies associated with decreased thrombopoiesis. Additionally, alcohol has a direct toxic effect on megakaryocytes, which contributes to prevailing vitamin deficiencies and decreased bone marrow production of platelets. Treatment of bleeding in liver failure with vitamin K usually is not successful given the lack of hepatic synthetic function. Whole blood both corrects the red blood cell deficit and is as effective as fresh frozen plasma in correcting coagulation factor deficits if the units of blood have not been banked for an extended period of time (14,61). Platelet transfusions should be judiciously used to raise the platelet count to above 100 × 109/L if bleeding is encountered in this setting, and FFP should be provided to correct deficits in plasmatic coagulation factors (10,11,12,14).
Transfusion-Induced Bleeding
Major trauma, major orthopedic (spine, hip, or pelvis) or hepatic procedures (major hepatic resections, liver transplantation), or other causes of potentially life-threatening, exsanguinating hemorrhage is often associated with the need for what has been termed massive blood transfusion (21,22,61). This term has been variously defined but generally refers to administering the equivalent of one total blood volume or more to a patient in less than a 24-hour period. Due to consumption of coagulation factors and platelets, release of inflammatory mediators, dilution of elements necessary for the optimal function of the coagulation cascade, hypocalcemia, hypothermia, fibrinolysis, and alterations in acid-base homeostasis, a coagulopathy often develops in this scenario (21,22,61,64). Banked blood is a negligible source of viable platelets, which rapidly deteriorate under conditions of cold storage (10,12,14). Additionally, depending on the age of the unit, plasmatic coagulation factors may also be diminished in activity. The derangement in clotting represents nonlocalized, nonsurgical bleeding that is characterized by sanguinous oozing from all raw surfaces, including any wounds, mucosal or peritoneal surfaces, and percutaneous entry sites. Numerous risk factors for the development of this condition (particularly in the setting of trauma) include high injury severity score, acidosis, hypothermia, and hypotension (2,4,20,21). Prophylactic administration of platelets during massive blood transfusion in an attempt to prevent the development of a coagulopathy has been advocated by many centers, but the efficacy of this policy is unproven. Studies have failed to demonstrate conclusively a benefit to this approach (10,12,14,22,61,65,66,67). It is reasonable, however, to administer platelets to a bleeding patient or one with DIC if the platelet count is less than 50 × 109/L. For those patients with rapid bleeding or with multiple traumatic injuries undergoing surgery or other high-risk procedures, a goal of at least 100 × 109/L has been recommended, albeit in the absence of high-level evidence-based data. One unit of platelet concentrate typically raises the platelet count by 5 × 109/L, whereas one apheresis concentrate raises the platelet count by 20 to 25 × 109/L in an average 70-kg adult. Despite plasmatic coagulation factor activity deteriorating with storage, dilution of these factors below levels required for adequate hemostasis is rare. Therefore, routine administration of FFP after an arbitrary number of units of banked blood has been transfused is not supported by stringent investigations (2,4,10,11,14,67,68,69,70,71). If there is bleeding due to a coagulopathy with concomitant prolongation of the PT and PTT of more than 1.5 times normal, FFP should be infused (to normalize the PTT and achieve a PT international normalized ratio [INR] <1.5). There is emphasis on early correction since timid replacement of coagulation products will lead to more bleeding and being more coagulopathic, leading to a vicious cycle. Cryoprecipitate should be considered if the fibrinogen level is less than 0.8 g/dL. Recombinant activated factor VIIa has demonstrated promise in significantly reducing bleeding and blood transfusion requirements in patients with traumatic injuries, or surgical procedures attended by significant hemorrhage, in addition to liver transplantation (72,73,74,75,76,77,78).
Complications directly attributable to blood transfusions have been described earlier. However, in the setting of massive blood transfusion, numerous points are warranted. Hemolytic transfusion reactions may occur as a result of ABO incompatibility between the actual units of blood transfused, or potentially because of clerical error, magnified by the volume of units required in a relatively short period of time. Red blood cells are a rich source of tissue thromboplastin, and hemolysis results in a massive release of this extrinsic system activator. This manifests as a generalized oozing from all raw surfaces, similar to the coagulopathy seen with massive blood transfusions, rendering this a difficult diagnosis to make. Hemoglobinuria may be observed due to filtration of plasmafree hemoglobin into the urine. Consistent with other types of transfusion reactions, the coagulation abnormality corrects rapidly once the offending transfusion is terminated. Treatment must be directed toward prevention of hemoglobin casts precipitating in the acidic environment of the collecting tubules, resulting in acute tubular necrosis and acute renal failure. Appropriate fluid resuscitation to maintain intravascular volume and to initiate a diuresis is paramount. Mannitol, an osmotic diuretic and free radical scavenger, is administered intravenously as an adjunct to maintain urine flow, and intravenous sodium bicarbonate may be considered to alkalinize the urine to avoid further precipitation of hemoglobin in the renal tubules.
Cardiopulmonary Bypass
Post-operative bleeding is a frequent impediment of cardiopulmonary bypass. Numerous mechanisms are apparently involved, which include contact factor (factors XII and XI) activation, elevations of tissue plasminogen activator level and tissue thromboplastin, dilution of plasmatic coagulation factors, residual effects of systemic heparinization, hypothermia, platelet function defects, and failure of surgical hemostasis (23,24,25,26). Some investigations have demonstrated a 30% to 50% decrement in platelet count attributable to the shearing forces that are encountered in the bypass apparatus. The routine use of antiplatelet agents in patients with cardiac disease, such as aspirin and thienopyridine derivatives (clopidogrel or ticlopidine), also contributes to the increased risk for bleeding. The combined effects of aspirin and a thienopyridine derivative, such as clopidogrel, on bleeding complications is synergistic and not additive. In approximately 4% to 5% of patients, surgical re-exploration of the mediastinum is necessary, which varies based on the original procedure performed (23,25,79). Several criteria for re-exploration have been proposed that have in common the rate of blood loss from mediastinal or chest tubes. Criteria variably used include blood loss from chest or mediastinal tubes of 300 mL/hour within the first 3 hours; total blood loss of 1,000 mL after 4 hours; a sudden increase in bleeding (>300 mL/hour) in a patient who previously had minimal drainage; or evidence of cardiac tamponade. A coagulopathy must never be presumed to be the cause of bleeding postoperatively unless surgical causes of bleeding have first been excluded. A site of localized bleeding (surgical failure) is identified in more than 50% of patients re-explored based on these types of criteria (23,25,80).
Despite multiple contributing factors to the bleeding that occurs in these patients, the prime offender is collectively believed to be secondary to qualitative platelet function defects (1,23,24,25,26,79,80). The bypass circuitry results in platelet activation with degranulation and aggregation. Although this functional deficit is transient, increased time on bypass, hypothermia, and antiplatelet medications significantly exacerbates this condition. Laboratory analysis reveals a prolongation of the bleeding time with impaired adhesion and aggregation, particularly in the presence of adenosine diphosphate (ADP) and ristocetin. This latter finding is believed to be linked to low levels of vWF found in plasma after cardiopulmonary bypass. Hence, some investigators have proposed use of desmopressin in this circumstance (and in patients with a history of preoperative use of aspirin), to increase levels of vWF by stimulating release from endothelial cells, increasing the glycoprotein receptors on platelets, and increasing the level of factor VIII and tissue plasminogen activator. However, others believe that this practice increases the risk of graft thrombosis and coronary occlusion (2,79,80). Additionally, peer-reviewed, reported outcome data in this circumstance are indeterminate (2). Usually the acquired qualitative platelet function defect resolves within 4 hours of completion of cardiopulmonary bypass without any intervention. In instances where there is prolonged nonsurgical postoperative bleeding, platelet transfusions are often beneficial (10,14,80).
Many pharmacologic agents have proven efficacy in the management of nonsurgical bleeding, particularly in the post–cardiopulmonary bypass setting. Desmopressin has already been described. Antifibrinolytic agents constitute a heterogenous group of drugs with proven efficacy in cardiac surgery patients. Aprotinin, a bovine serine protease inhibitor, inhibits plasmin and has been shown to reduce the need for red blood cell transfusions in several randomized trials and reduces the need for reoperation for nonsurgical bleeding (1,2,4,26,80,81,82). There are several adverse side effects that limit its usefulness. Thromboembolic phenomena, renal insufficiency, and allergic reactions (probably due to its bovine origin) have all been reported. The occurrence of serious anaphylactic reactions has been the impetus for use of a test dose prior to full implementation of the agent (2,80). Aprotinin has been removed from FDA approval until further studies are done confirming safety. Epsilon aminocaproic acid (EACA) and tranexamic acid are lysine analogues that inhibit binding of plasmin to fibrin. EACA appears to have the weakest antifibrinolytic effect compared to tranexamic acid and aprotinin, but nevertheless has been used in the cardiopulmonary bypass patient with some success. Tranexamic acid has met with considerable success in the reduction of postoperative blood loss and the reduced need for red blood cell transfusion in cardiac surgery, total knee arthroplasty, transurethral prostate surgery, and in oral surgery procedures. Neither EACA nor tranexamic acid are associated with thrombotic complications or anaphylactic reactions. The most commonly reported adverse reactions include nausea, diarrhea, and orthostatic reactions (83). Considering efficacy, side effect profile, and lower cost, tranexamic acid has several advantages over aprotinin (26,80,81,82,83).
Monitoring for increased risks of hemorrhage commences during surgery (84,85,86). The activated clotting time, which measures the effect of heparin on fibrin clot formation, has been traditionally used for intraoperative management of anticoagulation therapy. Thromboelastography, an assay popularized in Europe, can be used to determine if there is a platelet function abnormality, a deficit in plasmatic coagulation factors, the presence of circulating anticoagulants, or fibrinolysins in patients undergoing cardiopulmonary bypass (4,16,85,86). This technique evaluates clot formation and continually evaluates clot firmness in an integrated manner and is amenable to point-of-care testing, hence its usefulness both intraoperatively and postoperatively. No specific specimen processing is required, and whole blood is used for testing. This technology is rapidly gaining attention as a valuable adjunct to managing complex coagulation disorders. It has been efficacious in decreasing blood transfusion requirements during cardiac surgery and in liver transplantation (87,88). The effect of hypothermia on a patient's coagulation profile can also be determined simply by adjusting the temperature of the apparatus to correspond with the patient's core body temperature. Several automated analyzers are available commercially.
Summary
Postoperative bleeding requires immediate recognition of early shock, resuscitation, and differentiation between surgical and nonsurgical bleed. Knowledge of coagulation and appropriate replacement of blood and blood products is essential. Careful evaluation of the patient to identify primary or secondary coagulation disorders is necessary.
References
1. Adams GL, Manson RJ, Turner I, et al. The balance of thrombosis and hemorrhage in surgery. Hematol Oncol Clin N Am. 2007;21:13.
2. Rossaint R, Cerny V, Coats TJ, et al. Key issues in advanced bleeding care in trauma. Shock. 2006;26:322.
3. Staudinger T, Locker GJ, Frass M. Management of acquired coagulation disorders in emergency and intensive care medicine. Semin Thromb Hemost. 1996;22:93.
4. Fries D, Innerhofer P, Schobersberger W. Coagulation management in trauma patients. Curr Opin Anaesthesiol. 2002;15:217.
5. Rohrer M, Natale A. Effect of hypothermia on the coagulation cascade. Crit Care Med. 1992;20:1402.
6. Dempfle C. Disseminated intravascular coagulation and coagulation disorders. Curr Opin Anaesthesiol. 2004;17:125.
7. Levi M, Ten Cate H. Disseminated intravascular coagulation. N Eng J Med. 1999;341:586.
8. Bick RL. Disseminated intravascular coagulation: current concepts of etiology, pathophysiology, diagnosis, and treatment. Hematol Oncol Clin North Am. 2003;17:149.
9. Zeerleder S, Hack CE, Wuillemin WA. Disseminated intravascular coagulation in sepsis. Chest. 2005;128:2864.
10. Practice Guidelines Development Task Force of the College of American Pathologists. Practice parameters for the use of fresh frozen plasma, cryoprecipitate, and platelets. JAMA. 1994;271:777.
11. British Committee for Standards in Haematology. Guidelines for the use of fresh frozen plasma. Transfusion Med. 1992;2:57.
12. British Committee for Standards in Haematology Blood Transfusion Task Force. Guidelines for the use of platelet transfusions. Br J Haematol. 2003;122:10.
13. Shortt J, Dunkley S, Rickard K, et al. Efficacy and safety of a high purity, double virus inactivated factor VIII/von Willebrand factor concentrate (Biostate) in patients with von Willebrand disorder requiring invasive or surgical procedures. Hemophilia. 2007;13:144.
14. American Society of Anesthesiologists Task Force on Blood Component Therapy: practice guidelines for blood component therapy. Anesthesiology. 1996;84:732.
15. Kearon C, Hirsh J. Management of anticoagulation before and after elective surgery. N Engl J Med. 1997;336:1506.
16. Shore-Lesserson L. Platelet inhibitors and monitoring platelet function: implications for bleeding. Hematol Oncol Clin North Am. 2007;21:51.
17. George JN, Woolf SH, Raskob GE, et al. Diagnosis and treatment of idiopathic thrombocytopenic purpura: recommendations of the American Society of Hematology. Ann Intern Med. 1997;126:319.
18. Senzolo M, Burra P, Cholongitas E, et al. New insights into the coagulopathy of liver disease and liver transplantation. World J Gastroenterol. 2006;12:7725.
19. Ansell J, Hirsh J, Poller L, et al. The pharmacology and management of the vitamin K antagonists: the seventh ACCP conference on antithrombotic and thrombolitic therapy. Chest. 2004;126:204.
20. MacLeod JB, Lynn M, McKenney MG, et al. Early coagulopathy predicts mortality in trauma. J Trauma. 2003;55:39.
21. Cosgrif N, Moore EE, Sauaia A, et al. Predicting life-threatening coagulopathy in the massively transfused trauma patient: hypothermia and acidosis revisited. J Trauma. 1997;42:857.
22. Levy JH. Massive transfusion coagulopathy. Semin Hematol. 2006;43:S59.
23. Despotis GJ, Filos KS, Zoys TN, et al. Factors associated with excessive postoperative blood loss and hemostatic transfusion requirements: a multivariate analysis in cardiac surgical patients. Anesth Analg. 1996;82:13.
24. Khuri SF, Wolfe JA, Josa M, et al. Hematologic changes during and after cardiopulmonary bypass and their relationship to the bleeding time and nonsurgical blood loss. J Thorac Cardiovasc Surg. 1992;104:94.
25. Dacey LJ, Munoz JJ, Baribeau YR, et al. Reexploration for hemorrhage following coronary artery bypass grafting: incidence and risk factors. Arch Surg. 1998;133:442.
26. Levi M, Cromheecke ME, de Jonge E, et al. Pharmacological strategies to decrease excessive blood loss in cardiac surgery: a meta-analysis of clinically relevant endpoints. Lancet. 1999;354:1940.
27. Dentali F, Ageno W, Crowther M. Treatment of coumarin-associated coagulopathy: a systematic review and proposed treatment algorithms. J Thromb Haemost. 2006;4:1853.
28. Martinez J, MacDonald KA, Palascak JE: The role of sialic acid in the dysfibrinogenemia associated with liver disease: distribution of sialic acid on the constituent chains. Blood. 1983;61:1196.
29. MacGillivray TE. Fibrin sealants and glues. J Card Surg. 2003;18:480.
30. Kunkel JM, Gomez ER, Spebar MJ, et al. Wound hematomas after carotid endarterectomy. Am J Surg. 1984;148:844.
31. Kwaan HC, Nabhan C. Hereditary and acquired defects in the fibrinolytic system associated with thrombosis. Hematol Oncol Clin N Am. 2003;17:103.
32. Wolberg AS. Thrombin generation and fibrin clot structure. Blood Rev. 2007;21:131.
33. Murphy S. Disorders of platelet production. In: Colman RW, Hirsch J, Marder VJ, et al., eds. Hemostasis and Thrombosis. Philadelphia, PA: JB Lippincott Co, 1982:259.
34. Jacobson MA, Young LS. New developments in the treatment of gram-negative bacteremia. West J Med. 1986;144:185.
35. George JN. Evaluation and management of patients with thrombotic thrombocytopenic purpura. J Intensive Care Med. 2007;22:82.
36. Franchini M. Thrombotic microangiopathies: an update. Hematology. 2006;11:139.
37. Harlan JM. Thrombocytopenia due to non-immune platelet destruction. Clin Haematol. 1983;12:39.
38. McCrae KR, Herman JH. Posttransfusion purpura: two unusual cases and a literature review. Am J Hematol. 1996;52:205.
39. Koene HR. Critical issues of current and future developments in the treatment of immune thrombocytopenic purpura. Pediatr Blood Cancer. 2006;47(5 Suppl):705.
40. Bussel J. Treatment of immune thrombocytopenic purpura in adults Semin Hematol 2006;43(3 Suppl 5):S3.
41. Warkentin T, Greinacher A. Heparin-induced thrombocytopenia: recognition, treatment, and prevention. The seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest. 2004;126:311.
42. Mielke CH Jr. Influence of aspirin on platelets and the bleeding time. Am J Med. 1985;74:72.
43. Valeri CR, Feingold H, Cassidy G, et al; Hypothermia-induced reversible platelet dysfunction. Ann Surg. 1987;205:175.
44. Escolar G, Diaz-Ricart M, Cases A. Uremic platelet dysfunction: past and present. Curr Hematol Rep. 2005;4:359.
45. Boccardo P, Remuzzi G, Galbusera M. Platelet dysfunction in renal failure. Semin Thromb Hemost. 2004;30:579.
46. Kaw D, Malhotra D. Platelet dysfunction and end-stage renal disease. Semin Dial. 2006;19:317.
47. Gentilello LM, Cortes V, Moujaes S, et al. Continuous arteriovenous rewarming: experimental results and thermodynamic model simulation of treatment for hypothermia. J Trauma. 1990;30:1436.
48. Galli M, Barbui T. Antiphospholipid syndrome: clinical and diagnostic utility of laboratory tests. Semin Thromb Hemost. 2005;31:17.
49. Galli M, Luciani D, Bertolini G, et al. Lupus anticoagulants are stronger risk factors for thrombosis than anticardiolipin antibodies in the antiphospholipid syndrome: a systematic review of the literature. Blood. 2003;101:1827.
50. Sultan Y. Acquired hemophilia and its treatment. Blood Coagul Fibrinolysis. 1997;8(Suppl 1):S15.
51. Franchini M. Acquired hemophilia A. Hematology. 2006;11:119.
52. Lusher JM. Inhibitor antibodies to factor VIII and factor IX: management. Semin Thromb Hemost. 2000;26:179.
53. Thiagarajan P, Niemetz J. Procoagulant-tissue factor activity of circulating peripheral blood leukocytes: results of in vivo studies. Thromb Res. 1980;17:891.
54. Maier RV, Ulevitch RJ. The induction of a unique procoagulant activity in rabbit hepatic macrophages by bacterial lipopolysaccharides. J Immunol. 1981;127:1596.
55. Schorer AE, Rick PD, Swaim WR, et al. Structural features of endotoxin required for stimulation of endothelial cell tissue factor production: exposure of preformed tissue factor after oxidant-mediated endothelial cell injury. J Lab Clin Med. 1985;106:38.
56. Bennett B, Towler HM. Haemostatic response to trauma. Br Med Bull. 1985;41:274.
57. Kaufman HH, Hui KS, Mattson JC, et al. Clinicopathological correlations of disseminated intravascular coagulation in patients with head injury. Neurosurgery. 1984;15:34.
58. Risberg B, Medegard A, Heideman M, et al. Early activation of humoral proteolytic systems in patients with multiple trauma. Crit Care Med. 1986;14:917.
59. Lasson A, Ohlsson K. Consumptive coagulopathy, fibrinolysis and protease-antiprotease interactions during acute human pancreatitis. Thromb Res. 1986;41:167.
60. Leeksma OC, Meijer-Huizinga F, Stoepman-van Dalen EA, et al. Fibrinopeptide A and the phosphate content of fibrinogen in venous thromboembolism and disseminated intravascular coagulation. Blood. 1986;67:1460.
61. Counts RB, Haisch C, Simon TL, et al. Hemostasis in massively transfused trauma patients. Ann Surg. 1979;190:91.
62. Feinstein DI: Diagnosis and management of disseminated intravascular coagulation: the role of heparin therapy. Blood. 1982;60:284.
63. Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of recombinant human activated protein c for severe sepsis. N Engl J Med. 2001;344:699.
64. Spahn DR. Hypocalcemia in trauma: frequent but frequently undetected and underestimated. Crit Care Med. 2005;33:2124.
65. Reed RLD, Ciavarella D, Heimbach DM, et al. Prophylactic platelet administration during massive transfusion: a prospective, randomized, double-blind clinical study. Ann Surg. 1986;203:40.
66. Segal JB, Dzik WH. Paucity of studies to support that abnormal coagulation test results predict bleeding in the setting of invasive procedures: an evidence-based review. Transfusion. 2005;45:1413.
67. Gajic O, Dzik WH, Toy P. Fresh frozen plasma and platelet transfusion for nonbleeding patients in the intensive care unit: benefit or harm? Crit Care Med. 2006;34(5 Suppl):S170.
68. Holland L, Sarode R. Should plasma be transfused prophylactically before invasive procedures? Curr Opin Hematol. 2006;13:447.
69. Dara SI, Rana R, Afessa B, et al. Fresh frozen plasma transfusion in critically ill medical patients with coagulopathy. Crit Care Med. 2005;33:2667.
70. Gonzalez EA, Moore FA, Holcomb JB, et al. Fresh frozen plasma should be given earlier to patients requiring massive transfusion. J Trauma. 2007;62:112.
71. Abdel-Wahab OI, Healy B, Dzik WH. Effect of fresh-frozen plasma transfusion on prothrombin time and bleeding in patients with mild coagulation abnormalities. Transfusion. 2006;46:1279.
72. Dutton RP, McCunn M, Hyder M, et al. Factor VIIa for correction of traumatic coagulopathy. J Trauma. 2004;57:709.
73. Gala B, Quintela J, Aguirrezabalaga J, et al. Benefits of recombinant activated factor VII in complicated liver transplantation. Transplantation Proc. 2005;37:3919.
74. Rossaint R, Boffard K, Warren B, et al. Decreased transfusion utilization using recombinant factor VIIa as an adjunct in trauma. Intensive Care Med. 2004;30(Suppl):S199.
75. Boffard KD, Riou B, Warren B, et al. Recombinant factor VIIa as adjunctive therapy for bleeding control in severely injured trauma patients: two parallel randomized, placebo-controlled, double-blind clinical trials. J Trauma. 2005;59:8.
76. Rizoli SB, Boffard KD, Riou B, et al. Recombinant activated factor VII as an adjunctive therapy for bleeding control in severe trauma patients with coagulopathy: subgroup analysis from two randomized trials. Crit Care. 2006;10:R178.
77. Filsoufi F, Castillo JG, Rahmanian PB, et al. Effective management of refractory postcardiotomy bleeding with the use of recombinant activated factor VII. Ann Thor Surg. 2006;82:1779.
78. Levi M, Peters M, Buller HR. Efficacy and safety of recombinant factor VIIa for treatment of severe bleeding: a systematic review. Crit Care Med. 2005;33:883.
79. Woodman RC, LA Harker: Bleeding complications associated with cardiopulmonary bypass. Blood. 1990;76:1680.
80. Royston D. Blood-sparing drugs: aprotinin, tranexamic acid, and epsilon-aminocaproic acid. Int Anesthesiol Clin. 1995;33:155.
81. Levy JH. Anti-inflammatory strategies and hemostatic agents: old drugs, new ideas. Hematol Oncol Clin N Am. 2007;21:89.
82. Harker LA. Bleeding after cardiopulmonary bypass [editorial]. N Engl J Med. 1986;314:1446.
83. Dunn CJ, Goa KL. Tranexamic acid: a review of its use in surgery and other indications. Drugs. 1999;57:1005.
84. Spiess BD, Davalle M. Coagulation monitoring in the surgical intensive care unit. Crit Care Clin. 1988;4:605.
85. Despotis GJ, Santoro SA, Spitznagel E, et al. Prospective evaluation and clinical utility of on-site monitoring of coagulation in patients undergoing cardiac operations. J Thorac Cardiovasc Surg. 1994;107:271.
86. Bracey AW, Grigore AM, Nussmeier NA. Impact of platelet testing on presurgical screening and implications for cardiac and noncardiac surgical procedures. Am J Card. 2006;98(Suppl 1):S25.
87. Shore-Lesserson L, Manspeitzer HE, Deprio M, et al. Thrombelastography-guided transfusion algorithm reduces transfusion in complex cardiac surgery. Anesth Analg. 1999;88:312.
88. Kang YG, Martin DJ, Marquez J. Intraoperative changes in blood coagulation and thromboelastographic monitoring in liver transplantation. Anesth Analg. 1995;64:888.