Version May 2024
INTRODUCTION — The use of any anticoagulant is associated with an increased risk of bleeding, and bleeding complications can be life-threatening. Management of bleeding in individuals who are receiving a direct oral anticoagulant (DOAC) can be challenging because routine coagulation tests cannot generally be used to determine the degree of anticoagulation, and some of the reversal agents are difficult to access and may be prothrombotic.
This topic discusses our approach to managing bleeding in patients receiving DOACs.
Separate topic reviews discuss the following:
DOAC-associated bleeding:
●Intracerebral hemorrhage (ICH) – (See "Reversal of anticoagulation in intracranial hemorrhage".)
●Heavy menstrual bleeding – (See "Abnormal uterine bleeding in nonpregnant reproductive-age patients: Management", section on 'Patients on anticoagulant therapy'.)
●Prevention of bleeding – (See "Risks and prevention of bleeding with oral anticoagulants".)
●Perioperative management – (See "Perioperative management of patients receiving anticoagulants".)
Use of DOACs:
●General considerations – (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects".)
●Atrial fibrillation – (See "Atrial fibrillation in adults: Use of oral anticoagulants".)
●Venous thromboembolism (VTE) – (See "Venous thromboembolism: Initiation of anticoagulation" and "Venous thromboembolism: Anticoagulation after initial management".)
DEFINITION OF TERMS
Direct oral anticoagulants (DOACs) — Direct oral anticoagulants (DOACs; also called non-vitamin K oral anticoagulants [NOACs]) are oral medications that directly inhibit a specific enzyme in the coagulation cascade.
Available agents include those that directly inhibit thrombin (factor IIa) or factor Xa. Additional specific inhibitors of factors XIa and XIIa are in development. The positions of these enzymes in the coagulation cascade are illustrated in the figure (figure 1) and discussed in detail separately. (See "Overview of hemostasis".)
●Dabigatran – Dabigatran (Pradaxa) is the only oral direct thrombin inhibitor available for clinical use.
Parenteral direct thrombin inhibitors include argatroban and bivalirudin; these agents are discussed separately. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Argatroban' and "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Parenteral direct thrombin inhibitors'.)
●Rivaroxaban, apixaban, edoxaban – Rivaroxaban (Xarelto), apixaban (Eliquis), and edoxaban (Lixiana, Savaysa) are oral direct factor Xa ("ten-a") inhibitors. Of note, the generic names for these agents include "Xa-ban" (rivaroxaban, apixaban, edoxaban). Betrixaban was a factor Xa inhibitor that was withdrawn from the market.
Other anticoagulants also inhibit factors Xa and thrombin, but their effects are indirect. The effects of heparins (which inhibit factor Xa and to a lesser extent thrombin and other serine proteases) and fondaparinux (which exclusively inhibits factor Xa) are mediated through antithrombin (AT). Warfarin reduces the synthesis of functional factors II (prothrombin), VII, IX, and X; its anticoagulant effect is due to reduced ability to generate active thrombin, rather than an inhibition of activated enzymes.
Bleeding severity — Bleeding is a spectrum, and bleeding that is initially minor can sometimes become more significant or even life-threatening. Thus, clinical judgment is required in all cases of bleeding to determine the risk to the patient and the appropriate degree of concern regarding whether the bleeding is resolving or worsening.
It is worth remembering that the initial presentation of bleeding may be misleading; what appears to be massive lower gastrointestinal bleeding may in retrospect be less serious than initially suspected, while many forms of "occult" bleeding may present innocuously but ultimately lead to major morbidity or death; retroperitoneal bleeding is an example.
Bleeding severity cannot initially be measured with hemoglobin determinations; development of anemia will be delayed until after fluid resuscitation or re-equilibration from body water, which may take several hours.
We define serious/major bleeding as bleeding that is associated with a risk of blood transfusion or bleeding into a critical closed space (eg, intracranial bleeding, compartment syndrome). Major bleeding may also include bleeding requiring an intervention for management (eg, surgery, interventional radiology procedures, endoscopic treatments). Major bleeding has a significant risk of immediate morbidity, regardless of the cause. Some major bleeding may be life-threatening and patients with major bleeding events are at high risk of death during the hospitalization prompted by that bleeding.
Minor but still clinically significant bleeding includes bleeding requiring a healthcare assessment or less invasive treatment, such as heavy menstrual bleeding, ecchymosis, or epistaxis. All types of bleeding are important to the patient; however, minor bleeding usually will not require interruption of anticoagulant therapy.
PATIENT ASSESSMENT — The appropriate management of DOAC-associated bleeding depends on risk stratification for the severity of the bleeding, anticoagulation status, and underlying indication for the anticoagulant.
Initial assessment — We initially assess the severity of bleeding and the degree of hemostatic impairment. This includes a thorough history and complete review of medications. We specifically address the following from the history:
●How severe is the bleeding and where is it located?
●Is the patient actively bleeding now?
●Which agent is the patient receiving?
●When was the last dose of anticoagulant administered (and what is the expected duration of effect)?
●Could the patient have taken an intentional or unintentional overdose of the anticoagulant?
●Does the patient have a history of renal or hepatic disease, or evidence of an underlying bleeding disorder, that might cause excessive anticoagulant effect in the setting of standard drug dosing?
●Is the patient taking other medication(s) that could affect hemostasis (eg, aspirin, clopidogrel, ginseng)?
●Does the patient have other comorbidities that could promote bleeding (eg, liver disease, uremia, thrombocytopenia)?
●What is the indication for anticoagulation and how great is the risk of thrombosis?
Assessment of bleeding — More severe bleeding will require more aggressive interventions than minor bleeding. Bleeding is assessed according to the site, rate of hemorrhage, and amount of blood loss. This includes a thorough physical examination with serial measurement of vital signs. (See 'Bleeding severity' above.)
Serial measurement of the hemoglobin level may be appropriate if there is concern about significant blood loss; of note, a fall in hemoglobin level is likely to be delayed with acute massive blood loss. (See 'Other limited laboratory testing' below.)
Imaging studies (eg, computed tomography scanning for intracranial or retroperitoneal hemorrhage) and/or endoscopy to visualize the bleeding site may also be required.
Many patients presenting with minor bleeding, slow blood loss, or simple anemia without evidence of bleeding may not require DOAC reversal. Given the short half-lives of these drugs, even short-term but unneeded interruption of therapy could result in avoidable thrombosis. Occasionally, bleeding that appears significant in fact is not (eg, some epistaxis or hemorrhoidal bleeds); in such cases, observation and local measures such as ice and pressure may allow resolution of the bleeding without requiring aggressive treatments that might expose the patient to the risks of reversal strategies and potential thrombosis due to withdrawal of anticoagulants.
Assessment of anticoagulation status — The degree of anticoagulation is important both for predicting the course of the bleeding episode and for determining which interventions will be required. The anticoagulation status depends on the specific agent, dose, time since the last dose, and renal (and to a lesser extent hepatic) function.
Interval since last dose — We consider anticoagulation to have resolved fully after five half-lives have elapsed since the last dose. We use the following half-lives for patients with normal renal function [1]:
●Dabigatran – 12 to 17 hours; five half-lives will have elapsed by day 2.5 to 3.5 after the last dose.
●Rivaroxaban – 5 to 9 hours; five half-lives will have elapsed by day 1 to 2 after the last dose.
●Apixaban – 8 to 15 hours; five half-lives will have elapsed by day 1.5 to 3 after the last dose.
●Edoxaban – 6 to 11 hours; five half-lives will have elapsed by day 1.3 to 2 after the last dose.
Renal and hepatic function — The half-lives of the DOACs are dependent on renal (and to a lesser extent hepatic) function; thus, patients with renal and/or severe hepatic impairment may have a greater degree and/or duration of anticoagulation than patients with normal renal and hepatic function for a given dose and/or schedule.
●Dabigatran – Excretion is approximately 80 to 85 percent renal.
●Rivaroxaban – Excretion is approximately 35 percent renal; severe hepatic impairment could result in bio-accumulation.
●Apixaban – Excretion is approximately 25 percent renal; severe hepatic impairment could result in bio-accumulation.
●Edoxaban – Excretion is approximately 35 percent renal; severe hepatic impairment could result in bio-accumulation.
Coagulation testing — Coagulation testing is not used for determining the anticoagulation status of a patient receiving a DOAC. Prolonged coagulation times can be helpful in determining residual anticoagulant effect (table 1), but normal coagulation testing cannot necessarily be used as evidence that the anticoagulant effect of a particular DOAC has resolved or to eliminate the need for aggressive interventions. Clinicians should check with their local laboratory to determine whether (and which) rapidly available assay can exclude the presence of a significant DOAC effect.
We perform the following coagulation testing:
●Prothrombin time/international normalized ratio (PT/INR)
●Activated partial thromboplastin time (aPTT)
●Thrombin clotting time (TT) in patients with suspected dabigatran effect
Some laboratories include a rapid turnaround anti-factor Xa and thrombin time (DOAC screen).
If the relevant laboratory has determined that the PT or aPTT (or a modified version of these tests) is reliably prolonged by even low concentrations (eg, 30 to 50 ng/mL) of the agent in question, a normal value may be reassuring; this requires consultation with laboratory personnel regarding institution-specific testing that has been validated for use in assessing DOAC effect [2].
Specialized testing has limited availability but may be of use if it is available with a rapid turnaround time from a laboratory that is familiar with such testing. This testing may include anti-factor Xa heparin level (useful as a guide to the presence of a direct factor Xa inhibitor), quantitative factor Xa inhibitor levels, and quantitative dabigatran levels (eg, as determined using a dilute thrombin time). These tests may be of particular use when use of a reversal agent is being considered. Impairment of holistic coagulation tests, such as the TEG, may indicate the presence of a DOAC, but there is no evidence that these tests are changed in a way that predicts the amount of drug present.
●Dabigatran – Although we treat patients with persistent bleeding in the setting of normal basic coagulation testing as if they remain anticoagulated, an exception is a patient receiving dabigatran who has a normal TT; we consider this sufficient to eliminate the possibility of continued dabigatran effect. However, the TT is extremely sensitive to the effects of dabigatran and can be prolonged even by trivial amounts of the drug. If available, a calibrated dilute TT is preferable since it correlates linearly with dabigatran concentration.
●Rivaroxaban, apixaban, edoxaban – Specific assays for anti-factor Xa activity are available in many institutions, although only a small number of clinicians will have access to these assays in a timely fashion.
•The absence of anti-factor Xa activity, regardless of how an assay has been calibrated, indicates that no clinically relevant anti-factor Xa drug effect is present [2].
•Increased anti-factor Xa activity may reflect the presence of continued anti-factor Xa anticoagulant effect; however, unless the assay used has been calibrated for the specific anticoagulant the patient is taking, the amount of anticoagulant effect present cannot be reliably determined. Thus, it is important to consult with laboratory personnel regarding how the assay behaves in the presence of each of the different factor Xa inhibitors.
We do not use thromboelastography (TEG) or related point-of-care coagulation tests to manage DOAC-associated bleeding. DOACs have been demonstrated to affect the results of TEG and similar tests; however, data correlating these tests with drug concentration, clinical bleeding, and/or anticoagulant effect are lacking [3-7].
Patients receiving a DOAC may also have coagulation abnormalities unrelated to the DOAC, such as vitamin K deficiency, dilution coagulopathy, or disseminated intravascular coagulation (DIC) due to trauma or infection; this may also require investigation. (See 'Other limited laboratory testing' below.)
Other limited laboratory testing — No additional laboratory testing is absolutely required beyond coagulation testing outlined above. (See 'Assessment of anticoagulation status' above.)
However, depending on the severity of bleeding, we often perform a complete blood count (CBC) to assess the serum hemoglobin level and platelet count because transfusions of red blood cells and/or platelets may be indicated in some patients.
It may also be appropriate in some patients (eg, trauma, incomplete history) to evaluate for other causes of prolonged coagulation times, including liver disease, nutritional depletion of vitamin K, disseminated intravascular coagulation (DIC), or the use of a non-DOAC anticoagulant. Assessment of abnormal renal or hepatic function is important because these conditions could affect bleeding directly (by causing impaired platelet function or coagulopathy, respectively) or indirectly (by reducing metabolism of the anticoagulant).
●Hemoglobin level – The hemoglobin level is helpful in cases of major bleeding (eg, gastrointestinal), both to assess the severity of bleeding and to determine whether red blood cell transfusion is indicated. Patients with active bleeding cannot be managed exclusively by serial hemoglobin levels, because the rate of bleeding may outpace volume redistribution and/or the ability to accurately measure changes in blood loss. (See "Use of blood products in the critically ill", section on 'Red blood cells' and "Approach to acute upper gastrointestinal bleeding in adults", section on 'Blood product transfusions' and "Initial management of moderate to severe hemorrhage in the adult trauma patient".)
Hemoglobin thresholds for transfusion in hemodynamically stable patients are discussed separately. (See "Indications and hemoglobin thresholds for RBC transfusion in adults".)
●Platelet count – We usually measure the platelet count to eliminate the possibility of thrombocytopenia contributing to bleeding. Platelet transfusions are indicated for the treatment of major bleeding in patients with thrombocytopenia regardless of the underlying cause. Specific thresholds for platelet transfusion depend on the severity of the bleeding. (See "Platelet transfusion: Indications, ordering, and associated risks", section on 'Indications for platelet transfusion'.)
The evaluation of unexpected thrombocytopenia is discussed separately. (See "Diagnostic approach to thrombocytopenia in adults".)
The platelet count is normal in patients receiving aspirin and other antiplatelet agents, but platelet transfusion may be helpful if bleeding is especially severe.
●Renal function tests – Given the renal dependence of DOACs, determination of creatinine and calculation of the creatinine clearance should be performed in all patients with significant, unexplained bleeding. Uremia may also impair platelet function, further reducing hemostasis. (See "Uremic platelet dysfunction".)
●Liver function tests – We assess liver synthetic function in patients with prolonged PT or other signs or symptoms suggestive of liver disease, to assess the possibilities of reduced clearance of direct factor Xa inhibitors, which are partially metabolized in the liver. Diminished hepatic function may also contribute to bleeding if coagulation factor synthesis is impaired. These issues are discussed separately. (See "Clinical use of coagulation tests", section on 'Evaluation of abnormal results'.)
●Tests for DIC – Patients with suspected disseminated intravascular coagulation (DIC) in the setting of trauma or sepsis may have additional measurements such as fibrinogen and D-dimer testing. This subject is discussed in detail separately. (See "Evaluation and management of disseminated intravascular coagulation (DIC) in adults", section on 'Diagnostic evaluation'.)
MAJOR BLEEDING
Overview of management — Patients with serious/major bleeding should be managed in an intensive care setting with appropriate hemodynamic support. Options for the management of bleeding include observation; drug removal with activated charcoal and/or hemodialysis; and active interventions including administration of an antifibrinolytic agent, clotting factor products (eg, prothrombin complex concentrates, activated prothrombin complex concentrates), specific reversal agents (eg, idarucizumab for dabigatran, andexanet alfa for direct factor Xa inhibitors), and/or surgery [8-10].
The appropriate approach from among these options depends on ongoing assessment of the severity of bleeding (table 2). These therapies are discussed in more detail below. (See 'Anticoagulant reversal' below.)
Additional aspects of management for all patients with major bleeding include the following:
●Immediate discontinuation of all anticoagulant and antiplatelet therapy must be clearly described in the medical record and ordering system. This includes discontinuation of the anticoagulant the patient is taking, and avoidance of other anticoagulants (eg, "routine" orders for heparin administration for venous thromboembolism prophylaxis).
●Rapid and continuous hemodynamic assessment.
●Establishment of an effective airway and large-bore intravenous access.
●Optimization of body temperature, blood pH, and electrolyte balance, including calcium.
●Transfusions if required, including red blood cells for severe anemia or ongoing blood loss; platelets for thrombocytopenia and/or severe platelet dysfunction; and plasma for trauma-associated coagulopathy. (See 'Transfusions if needed' below.)
Early involvement of appropriate specialists is vital for definitive interventions at the site(s) of bleeding, which can be both diagnostic and therapeutic. The consulting endoscopist, interventional radiologist, or surgeon should be notified immediately regarding the potential need for intervention in a patient with bleeding that is (or may become) severe.
●Intracerebral hemorrhage (ICH) – ICH is of great concern because bleeding increases intracranial pressure, and expansion of the hematoma is common. Importantly, the diagnosis of ICH is confirmed by noncontrast computed tomography (CT) or magnetic resonance imaging (MRI), whichever is fastest; however, empiric treatment for suspected ICH in the absence of intracranial imaging is appropriate in some cases (eg, when the time delay in obtaining intracranial imaging could be life-threatening).
Management is discussed separately. (See "Reversal of anticoagulation in intracranial hemorrhage".)
●Other bleeding sites – Other major bleeding sites may include retroperitoneal bleeding, compartment syndrome, and massive gastrointestinal bleeding. Additional considerations in the management of these types of major bleeding are discussed in detail separately. (See "Approach to acute upper gastrointestinal bleeding in adults" and "Approach to acute lower gastrointestinal bleeding in adults" and "Abnormal uterine bleeding in nonpregnant reproductive-age patients: Terminology, evaluation, and approach to diagnosis" and "Initial management of trauma in adults" and "Abdominal compartment syndrome in adults" and "Acute compartment syndrome of the extremities".)
Anticoagulant reversal
General strategy for anticoagulant reversal — Reversal of anticoagulation is generally thought to be desirable in a patient with serious or life-threatening bleeding who remains actively anticoagulated with a DOAC, although randomized trial data to support this practice are lacking. Available strategies for reversing the anticoagulant effect of DOAC agents include the following (table 2):
●A specific reversal agent/antidote (for dabigatran, idarucizumab; for the oral direct factor Xa inhibitors, andexanet alfa) (table 3)
●Nonspecific agents such as prothrombin complex concentrates (PCCs)
●Antifibrinolytic agents
●Desmopressin (DDAVP)
●Drug removal from the circulation and/or gastrointestinal tract
High-quality evidence from randomized trials is lacking for these strategies; thus, our practice is based on clinical experience and data from case series [8,11]. Our practice attempts to balance the risk of life-threatening bleeding with the risk of thrombosis, which may be increased based on the underlying condition for which anticoagulation was prescribed and/or prothrombotic effects of some therapies (eg, andexanet alfa, PCCs). This balance depends on the assessment of bleeding severity (eg, life-threatening versus major versus clinically significant but non-major) and such assessments are inherently subjective requiring clinical expertise and close examination of the patient by the treating clinician; input from other specialists (eg, neurologist, gastroenterologist, hematologist) may be especially helpful. (See 'Bleeding severity' above.)
In most cases of major (life-threatening or uncontrolled) DOAC-associated bleeding, we suggest the use of a pro-hemostatic agent (eg, PCC or a specific antidote [if available]), an antifibrinolytic agent (for all anticoagulants), and/or oral activated charcoal (for all recently ingested anticoagulants). (See 'Dabigatran reversal' below and 'Factor Xa inhibitors' below.)
●PCC or antidote – We reserve PCCs and specific antidotes for the most serious/life-threatening cases of DOAC-associated bleeding when there is high confidence that the anticoagulant effect is present (based on knowing when the last dose was taken) and a specific antidote is not available or ineffective, such as ongoing bleeding that is likely to lead to death or permanent disability if not stopped immediately (table 3).
PCCs and some antidotes may have the potential to cause thrombosis at doses used to treat bleeding, and their impact on patient-important outcomes is not well established. Available data consist largely of non-bleeding volunteers, case reports, and preclinical models [12-21]. Thus, we avoid pro-hemostatic agents in patients with less severe bleeding. Importantly, use of these products should not be considered routine or "standard of care," and the clinician should be aware of the potential to precipitate thrombosis. (See 'Clotting factor products' below.)
●Therapies we avoid – We generally avoid the use of recombinant activated factor VII (rFVIIa), plasma, or cryoprecipitate in DOAC-associated bleeding because there are no data to support the use of these products, and there are associated risks (eg, thrombosis, volume overload, transfusion reactions). An exception is the use of plasma, PCC, cryoprecipitate, or fibrinogen concentrate to correct a coexisting coagulopathy (eg, from trauma or other settings with decreased levels of clotting factors and/or fibrinogen due to consumption or dilution). (See 'Transfusions if needed' below.)
Abnormal coagulation testing is consistent with continued DOAC effect, but normal testing does not necessarily eliminate the possibility of clinically important concentrations of these agents. Therefore, with the exception of the thrombin time (TT) in patients with suspected dabigatran-associated bleeding, the results of coagulation tests, or their trends over time, do not meaningfully inform reversal of DOAC-associated bleeding. Management is based on the patient's presentation, the timing of the last DOAC dose (if known with certainty) and change in their status over time rather than on the results of coagulation testing in most cases. (See 'Assessment of anticoagulation status' above.)
Dabigatran reversal — Idarucizumab (pronounced "I-dare-you-cizumab"; brand name Praxbind) is a humanized anti-dabigatran monoclonal antibody fragment that can be used for emergency reversal of the anticoagulant effect of dabigatran [22,23]. We would administer this agent, if available, for patients for whom more conservative bleeding management measures have been ineffective and who have life-threatening bleeding or for whom surgery is required on an urgent/emergency basis. We would only administer idarucizumab to patients with convincing evidence of significant dabigatran levels based on clinical history of ingestion or laboratory testing. Idarucizumab should not be administered to patients with a normal thrombin time (TT).
The dose is 5 grams (two 2.5 g vials), which can be administered either as two consecutive infusions or as a bolus (ie, injecting both vials consecutively via syringe).
The use of idarucizumab in reversing dabigatran anticoagulation was demonstrated in the RE-VERSE AD study (Reversal Effects of Idarucizumab on Active Dabigatran); the drug approval by the US Food and Drug Administration (FDA) in 2015 was based on an interim report of the first 90 patients on the study [24,25]. The final report described outcomes in 503 patients treated with idarucizumab to reverse the effect of dabigatran in the setting of an urgent procedure (202 patients) or bleeding (301 patients; the most common sites of bleeding were gastrointestinal, intracerebral, or traumatic). Most of the participants had atrial fibrillation and were receiving dabigatran for stroke prevention; the median age was 78 years. Outcomes were as follows [25]:
●Of 203 evaluable patients treated for bleeding, 134 (68 percent) had documented cessation of bleeding within 24 hours; the median time to assessment of hemostasis after administration of idarucizumab was 2.5 hours. Some individuals with intracerebral hemorrhage could not be evaluated since they did not have serial imaging studies.
●Of 197 evaluable patients undergoing an invasive procedure or surgery, hemostasis was judged to be normal in 184 (93 percent), mildly impaired in 10 (5 percent), and moderately impaired in 3 (2 percent). None had severely abnormal bleeding.
●Mortality was approximately 14 percent at one month and 19 percent at three months. Thrombotic events occurred in approximately 5 percent at one month and 7 percent at three months.
●Laboratory testing (dilute TT and/or ecarin clotting time [ECT]) showed that 92 percent of patients had some degree of anticoagulation prior to receiving idarucizumab. The anticoagulant effect was completely reversed (dilute TT or ECT normalized) within the first 15 minutes of treatment in almost all, and the reversal was maintained for 24 hours in most.
●Antibodies that reacted with idarucizumab were present in 19 participants before receiving the drug, and an additional nine developed antibodies during treatment. Titers were low, and these preexisting antibodies did not appear to interfere with the reversal therapy.
In a rat model, idarucizumab neutralized dabigatran activity at nanomolar concentrations and did not mimic thrombin functionally (ie, it did not bind to thrombin substrates, alter clotting parameters, or cause platelet aggregation on its own); thus, it is not expected to have any anticoagulant or prothrombotic effect [26].
Idarucizumab is a dabigatran-specific Fab fragment and has no known activity against direct factor Xa inhibitors, other thrombin (factor IIa) inhibitors, or other anticoagulants [26].
●Life-threatening bleeding – For patients who are at imminent risk of death from bleeding associated with dabigatran anticoagulation, we suggest administration of idarucizumab (table 3) [23]. We would not combine idarucizumab with "procoagulant" reversal agents such as activated prothrombin complex concentrate (aPCC).
If idarucizumab is not available, we suggest administering an aPCC such as factor eight inhibitor bypassing activity (FEIBA) at a fixed dose or a weight-based dose of 50 to 80 units/kg. If an aPCC is unavailable, an unactivated 4-factor or 3-factor PCC (table 4) would be a reasonable alternative at a fixed dose or a weight-based dose of 25 to 50 units/kg (see 'PCCs and aPCCs' below). The optimal weight-based dose is uncertain. Supplementation of 3-factor PCC with Fresh Frozen Plasma (FFP) has been used to supply factor VII, which is present at minimal levels in 3-factor PCCs. Importantly, treatment with aPCCs carries a real and substantial prothrombotic risk, and these agents have not been studied adequately in patients receiving dabigatran, so we avoid their use except in extreme clinical circumstances.
We also use the strategies listed below for less severe bleeding, including an antifibrinolytic agent, oral activated charcoal and/or hemodialysis when appropriate, and drug discontinuation.
●Major bleeding – For patients with major (including life-threatening) bleeding, we suggest administration of an antifibrinolytic agent (eg, tranexamic acid, epsilon-aminocaproic acid). The use of these agents may also be appropriate in individuals with less serious bleeding if the patient has ongoing bleeding or other comorbidities that increase bleeding risk. (See 'Antifibrinolytics and other pro-hemostatic therapies' below.)
We also suggest the administration of oral activated charcoal to remove the unabsorbed pro-drug (dabigatran etexilate) from the gastrointestinal tract if the last dose was within the previous two hours and the patient can tolerate the enteral administration (ie, not vomiting, adequate mental status, or via gastric tube if tracheally intubated). Oral activated charcoal may also be useful for individuals with less severe bleeding who may have taken an overdose of dabigatran, those whose bleeding is ongoing, or those who have other hemostatic abnormalities that could increase the risk of more severe bleeding. We advise consultation with a medical toxicologist or regional poison control center for a patient with an intentional excessive ingestion. A regional poison control center in the United States can be reached at 1-800-222-1222, and contact information for poison centers around the world is provided separately. (See "Gastrointestinal decontamination of the poisoned patient", section on 'Activated charcoal' and "Society guideline links: Regional poison control centers".)
Hemodialysis may also be used to remove active dabigatran from the circulation [27,28]. Studies in patients with renal impairment have shown removal of approximately one-half of the circulating drug by hemodialysis, with a minor redistribution after completion of the dialysis [29-32].
As noted above, discontinuation of dabigatran should be clearly described in the medical record and ordering system, and other anticoagulants should be avoided. (See 'Overview of management' above.)
●Minor bleeding – Minor bleeding can usually be managed conservatively. (See 'Minor bleeding' below.)
Factor Xa inhibitors
Overview of factor Xa inhibitor reversal — Available factor Xa inhibitors include apixaban, edoxaban and rivaroxaban. A summary of our overall approach is as follows, although we note that evidence from randomized trials is lacking regarding these strategies:
●Life-threatening bleeding – For patients who are at imminent risk of death from bleeding associated with direct factor Xa inhibitor anticoagulation, we suggest andexanet alfa or an unactivated 4-factor PCC (table 3).
•Dosing for andexanet, at one of two dose levels based on dose and timing of the factor Xa inhibitor, is described below. (See 'Andexanet alfa' below.)
•Dosing for unactivated 4 factor PCC can use a fixed dose of 2000 units or a weight-based dose of 25 to 50 units/kg, as discussed below. (See 'PCCs and aPCCs' below.)
There is insufficient evidence about risks and benefits to strongly favor either 4-factor PCC or andexanet over the other; these reversal agents have not been compared with each other directly in a randomized trial. If both andexanet and a 4-factor PCC are unavailable, a 3-factor PCC can be used (table 4); supplementation of 3-factor PCC with FFP has been used to supply factor VII, which is present at minimal levels in 3-factor PCCs. Evidence regarding the efficacy and safety of andexanet and PCCs is discussed below. Importantly, it is possible that treatment with andexanet or PCCs carries a real prothrombotic risk, so we avoid their use except in extreme clinical circumstances.
In addition to andexanet or a PCC, we also treat serious or life-threatening bleeding with an antifibrinolytic agent, oral activated charcoal when appropriate, and drug discontinuation.
●Major bleeding – For patients with major bleeding (including life-threatening bleeding), we suggest administration of an antifibrinolytic agent (eg, tranexamic acid, epsilon-aminocaproic acid). The use of these agents may also be appropriate in individuals with less serious bleeding if the patient has ongoing bleeding or other comorbidities that increase bleeding risk. (See 'Antifibrinolytics and other pro-hemostatic therapies' below.)
For most patients with bleeding where there is not an imminent risk of death, we suggest not using andexanet or PCC, based on the lack of high-quality evidence that the benefits will outweigh the risks. (See 'PCCs and aPCCs' below.)
For patients with major bleeding (including life-threatening bleeding), we suggest administration of activated charcoal to remove unabsorbed drug from the gastrointestinal tract if the last dose of the anticoagulant was within the previous two hours and the patient can tolerate enteral administration (ie, not vomiting, adequate mental status, or via gastric tube if tracheally intubated) [33-35]. In a patient with major bleeding following an intentional excessive ingestion within the previous six hours, we suggest charcoal administration since prolonged absorption has been reported (possibly from concretion formation or saturation of intestinal transporters) [36]. We advise consultation with a medical toxicologist or regional poison control center for a patient with an intentional ingestion. A regional poison control center in the United States can be reached at 1-800-222-1222, and contact information for poison centers around the world is provided separately. (See "Gastrointestinal decontamination of the poisoned patient", section on 'Activated charcoal' and "Society guideline links: Regional poison control centers".)
The available oral direct factor Xa inhibitors cannot be dialyzed, because they are highly protein-bound [37]. Charcoal hemofiltration has been suggested but not to our knowledge evaluated or used in patients with bleeding complications.
As noted above, discontinuation of the factor Xa inhibitor should be clearly described in the medical record and ordering system, and other anticoagulants should be avoided. (See 'Overview of management' above.)
●Minor bleeding – Minor bleeding can usually be managed conservatively. (See 'Minor bleeding' below.)
Andexanet alfa — Andexanet alfa (andexanet; coagulation factor Xa [recombinant], inactivated-zhzo, brand name AndexXa; previously called PRT064445) is a reversal agent for factor Xa inhibitors. It is a recombinantly produced, catalytically inactive form of factor Xa that acts as a "decoy" to bind and sequester the anticoagulant. Andexanet was approved by the US FDA in May of 2018 for the reversal of anticoagulation by rivaroxaban and apixaban in individuals with life-threatening or uncontrolled bleeding associated with these drugs [38].
We would use andexanet, when available, for patients who are likely to have significant amounts of Xa inhibitor “on board’, and who have life-threatening bleeding for whom more conservative bleeding management measures have been ineffective or for whom surgery is required on a truly urgent/emergency basis. 4-factor PCC is a reasonable alternative; these agents have not been compared directly in a randomized trial. We would not co-administer andexanet with a PCC product (ie, we would give one or the other, but not both).
For most patients with bleeding where there is not an imminent risk of death, we suggest not using andexanet (or a 4-factor PCC) based on the lack of high-quality evidence that benefits will outweigh the risks. We would only administer andexanet (or 4-factor PCC) to patients with convincing evidence of significant factor Xa inhibitor levels based on clinical history of ingestion or laboratory testing. In principle, this agent could reverse the effects of all of the direct factor Xa inhibitors as well as low molecular weight (LMW) heparins and fondaparinux, which inhibit factor Xa indirectly, but data on these other uses are lacking [39].
There are two dose levels:
●Low dose – A bolus of 400 mg given at 30 mg/minute, followed by an infusion of 480 mg given at 4 mg/minute for up to 120 minutes
●High dose – A bolus of 800 mg given at 30 mg/minute, followed by an infusion at 960 mg given at 8 mg/minute for up to 120 minutes
The low dose is used in patients who received a lower dose of factor Xa inhibitor (eg, rivaroxaban ≤10 mg, apixaban ≤5 mg) OR if ≥8 hours have elapsed since the last dose of factor Xa inhibitor; the higher dose is used for those who received a higher dose of factor Xa inhibitor (eg, rivaroxaban >10 mg, apixaban >5 mg, or dose unknown) within the previous 8 hours.
Evidence for the efficacy of andexanet in reversing direct factor Xa inhibitor anticoagulation comes from the single-arm ANNEXA-4 study [40,41]. In ANNEXA-4, andexanet was used to treat acute major bleeding in 352 patients (mean age, 77 years) who were receiving a factor Xa inhibitor for atrial fibrillation (representing 80 percent of the cohort) or venous thromboembolism (VTE) and had an anti-factor Xa activity above an arbitrarily defined threshold thought to indicate enhanced bleeding risk. Most of the patients were receiving rivaroxaban or apixaban. The dose level used was predicated on the type of factor Xa inhibitor and interval since administration, as described above.
●Hemostasis was judged to be excellent in 171 of 249 evaluable patients (69 percent) and good in 33 of 249 (13 percent, for a total of 82 percent good to excellent). While this rate of hemostatic efficacy compares favorably with that observed with PCC to reverse warfarin (82 percent with andexanet versus 72 percent with PCC using similar criteria), the degree to which either andexanet or PCC improve outcomes compared with a more conservative approach is not known [42]. Subset analysis for the site of bleeding found hemostasis to be good to excellent for 85 percent with gastrointestinal bleeding and 80 percent with intracranial bleeding. Over one-half of the patients (62 percent) restarted anticoagulation after recovering from bleeding.
●Reduction in anti-factor Xa activity was 92 percent for apixaban and rivaroxaban, and 75 percent for enoxaparin. Anti-factor Xa activity increased somewhat at the four- and eight-hour time-points but remained low (approximately 60 to 70 percent reduced). The correlation between anti-factor Xa activity reduction and achievement of hemostasis, as determined by blinded adjudicators using pre-specified criteria, was not strong.
●Adverse events included thromboses in 34 of 352 patients (10 percent), distributed across the 30 days of follow-up. Only 11 (3 percent) were within 5 days after receiving andexanet, suggesting that at least one-half of the thromboses could be related to the underlying thrombotic risk rather than the reversal drug. Thromboses included arterial and venous events (stroke, myocardial infarction, DVT, pulmonary embolism). Mortality within 30 days was 14 percent, a proportion comparable to all-cause mortality rates in other studies of patients with major anticoagulant-associated bleeding. Studies of andexanet in healthy volunteers did not observe any thromboses or other major adverse effects [43,44].
A randomized trial comparing andexanet with usual care in individuals with intracranial hemorrhage is ongoing [45].
As with PCC, andexanet has the potential to cause thrombosis, which provides the rationale for only using it in cases of serious or life-threatening bleeding. The package insert for andexanet includes a Boxed Warning regarding risks of arterial and venous thromboembolic and ischemic events including myocardial infarction, ischemic stroke, cardiac arrest, and sudden death [38].
Clotting factor products — Several products consisting of coagulation factors in the unactivated or activated state are available. These products have been developed for treating bleeding in other settings; data are sparse regarding the efficacy of any of these products in patients with DOAC-associated bleeding.
PCCs and aPCCs — Unactivated prothrombin complex concentrates (PCCs) and activated PCCs (aPCCs) both contain clotting factors purified from human plasma. They differ in their composition (eg, 3-factor PCCs contain factors II, IX, and X; 4-factor PCCs contain factors II, IX, X, and VII) and in whether the coagulation factors are in their inactive/zymogen form or whether a fraction of the clotting factors have been activated by proteolytic cleavage to the active enzyme (PCCs contain unactivated factors; aPCCs contain factor VIIa [factor VII in the activated form]). (See "Plasma derivatives and recombinant DNA-produced coagulation factors".)
Several mechanisms may contribute to the potential efficacy of these products; none have been definitively demonstrated to work by these mechanisms in the setting of DOAC-associated bleeding. Clotting factors circulate in an unactivated form and become activated at the site of vascular injury. Thus, administration of the unactivated products might be effective primarily by "overwhelming" the anticoagulant rather than by promoting hemostasis; however, this argument is hard to rationalize given that DOACs do not bind to zymogen (unactivated) clotting factors. At the site of bleeding, there is vascular injury and local thrombin generation; thus, the unactivated PCCs will become activated and enhance the local pro-hemostatic effect. By contrast, the activated products may promote hemostasis directly by providing much higher levels of activated factors than normal at sites of bleeding, which could theoretically overwhelm the local inhibitory effect of the DOAC.
Thrombosis is a potential risk with any pro-hemostatic agent. We generally regard aPCC as having a higher likelihood of precipitating a thrombotic event than unactivated PCC. However, there is little or no high-quality evidence that can be used to estimate the risk of thrombosis after the use of different PCC products, and it is not possible to quantify the degree to which an aPCC increases thrombosis risk (or promotes hemostasis) compared with an unactivated PCC.
As with andexanet, the evidence that PCCs will improve the outcome of a patient with DOAC-associated bleeding is weak and the risk of thrombosis is challenging to quantify. Thus, we would use one of these products only in life-threatening circumstances. For factor Xa-inhibitor reversal using a PCC, we have a weak preference for an unactivated PCC rather than an activated PCC because, in our opinion, it is less likely to cause harm (eg, it is less likely to cause a thrombosis). For dabigatran-associated bleeding, if we use a PCC, we have a weak preference for using an aPCC based on a theoretical role in bypassing thrombin and very limited clinical data, as discussed below.
●Unactivated prothrombin complex concentrates (PCCs) – Unactivated prothrombin complex concentrates (PCCs; also called "non-activated PCCs") are concentrates of coagulation factors and anticoagulants purified from plasma. They contain high levels of three or four coagulation factors (II, IX, and X in 3 factor PCCs; II, VII, IX, and X in 4 factor PCCs), along with protein C and S (table 4). Additional details regarding these products are presented separately. (See "Management of warfarin-associated bleeding or supratherapeutic INR", section on 'PCC products'.)
Unactivated PCCs are best studied for reversing warfarin anticoagulation; data regarding their efficacy in DOAC-associated bleeding are limited to observational studies, studies in healthy volunteers, and animal models [12-21,46-50]. A systematic review of case series published in 2019 identified 340 patients who received an unactivated PCC for direct factor Xa inhibitor-associated major bleeding (mostly rivaroxaban and apixaban) [51]. There have been a few studies in healthy volunteers but no randomized trials comparing PCC with other agents for serious bleeding associated with a factor Xa inhibitor. This systematic review found that the proportion of patients with overall successful bleeding management with PCC was 0.69 or 0.77, depending on the criteria used. The mortality rate was 0.16 and the rate of thrombosis was 0.04. The authors concluded that it was difficult to determine whether addition of PCC to anticoagulant discontinuation improved outcomes relative to anticoagulant discontinuation alone. These observations support a treatment plan where, if PCCs are administered, they are only administered to patients with life-threatening bleeding or bleeding likely to be associated with significant long-term morbidity. (See 'Factor Xa inhibitors' above.)
The optimal dosing for PCCs is uncertain. Fixed dose and weight-based dosing have not been compared in a randomized trial. There has been a general trend towards the use of fixed doses rather than weight-based dosing; this is primarily based on ease of use rather than evidence from clinical trials. Observational studies have shown fixed dosing to be effective [52]. Institutional protocols should be followed if available.
In the very rare instances where we choose to prescribe an unactivated PCC for imminently life-threatening bleeding associated with factor Xa inhibitor use, we would prescribe a fixed dose of 2000 units or a weight-based dose of 50 units/kg of actual body weight, irrespective of the patient's size. Some reports have used doses as low as 25 units/kg. Units refer to international units of factor IX activity. The use of fixed dose versus weight-based dose is individualized based on hospital protocols and the clinical scenario.
If the therapy appears ineffective, we typically would not re-administer the product, because it is unlikely that the risk-benefit ratio of doing so would be favorable. If hemostasis is improved after PCC administration, we also would not re-administer the medication unless a patient's renal function is impaired and the expected clearance of the DOAC will be significantly delayed. In such cases, a repeat dose of the PCC may be appropriate 12 to 24 hours after the first dose.
●Activated PCCs (aPCCs; factor VIII inhibitor activity bypassing agent [FEIBA]) – Activated PCCs (aPCCs) are PCCs that contain at least one factor in the activated form; these products have been developed and used for patients with hemophilia who have bleeding in the setting of a factor inhibitor. FEIBA is the only aPCC available in the United States; other aPCCs are available elsewhere. In FEIBA, factor VII is mostly activated.
When we use FEIBA, we start at a dose of 50 units/kg. If the therapy appears ineffective, we typically would not re-administer the product, because it is unlikely that the risk-benefit ratio of doing so would be favorable. If hemostasis is improved after aPCC administration, we also would not re-administer the medication unless a patient's renal function is impaired and the expected clearance of the DOAC will be significantly delayed. In such cases, a repeat dose of FEIBA may be appropriate 12 hours after the first dose; the total daily dose should not exceed 200 units/kg.
There are no high-quality data to support the use of aPCC in DOAC-associated bleeding or to inform the risk-benefit calculation for different dosing strategies; however, their use has been described in case reports, small series, and pre-clinical research settings. As examples:
•In a series of six patients with a DOAC-associated intracerebral hemorrhage (ICH) who were treated with FEIBA at 50 units/kg, none of the patients had expansion of the ICH, and there were no thrombotic or hemorrhagic complications [53].
•In a study in 10 non-bleeding individuals, FEIBA corrected rivaroxaban-induced impaired thrombin generation in a dose-dependent manner [12].
•In a pig polytrauma model, aPCC at a dose of 50 mg/kg prevented fatal bleeding when compared with placebo or lower dose aPCC [54].
•In baboons treated with rivaroxaban, FEIBA at a dose of 50 to 100 units/kg was reported to correct the bleeding time [55].
The extent to which the preclinical evidence correlates with in vivo efficacy for the management of patients with DOAC-associated bleeding is not known.
Unactivated PCCs and aPCCs are potentially prothrombotic, and clinical judgment is required in determining the individual bleeding and thrombotic risks of each patient on a case-by-case basis. The US FDA-approved product information for the use of unactivated PCCs to reverse warfarin effect in warfarin-associated bleeding lists a maximum dose of 5000 units [56]. This dose cap is reasonable for warfarin because the prothrombin time and international normalized ratio (PT and INR) can be measured following administration, and additional doses can be administered if repeat INR testing suggests that the warfarin effect has been incompletely reversed.
By contrast, the use of unactivated PCCs to treat factor Xa inhibitor-associated bleeding is off-label, and there is no measure such as the PT/INR that can be used to determine the need for further doses and no high-quality data to inform risk-benefit calculations for different doses. The product information notwithstanding, we are not aware of any clinical evidence or strong biologic rationale that would support a dose cap for PCC.
Recombinant activated factor VII (rFVIIa) — We generally avoid the use of recombinant activated factor VII (rFVIIa) to treat DOAC-associated bleeding. Although rFVIIa appears to restore some in vitro measures of coagulation, animal bleeding models do not suggest that rFVIIa would be beneficial for the treatment of dabigatran- or rivaroxaban-associated bleeding.
Plasma products — We do not use plasma products such as Fresh Frozen Plasma (FFP), Plasma Frozen Within 24 Hours After Phlebotomy (PF24), or cryoprecipitate to reverse the anticoagulant effect of DOACs. This practice is based on lack of any evidence that these products improve clinical outcomes, along with several potential risks including volume overload, transmission of infection, and transfusion reactions. (See "Clinical use of plasma components", section on 'Risks'.)
However, plasma products may be appropriate as part of a massive transfusion protocol in patients with severe bleeding who have developed a dilutional coagulopathy. (See 'Transfusions if needed' below and "Massive blood transfusion".)
Antifibrinolytics and other pro-hemostatic therapies — Other therapies to enhance hemostasis in a bleeding patient include the following:
●Antifibrinolytic agents including tranexamic acid and epsilon-aminocaproic acid can be used for severe bleeding. Both can be administered orally or intravenously; intravenous administration is generally favored for those with major or life-threatening bleeding.
•The usual oral dose of tranexamic acid is 1 to 1.5 grams every 8 to 12 hours for the duration of bleeding. Intravenous doses are less well characterized; doses of 10 to 20 mg/kg as an intravenous bolus followed by 10 mg per kg intravenously every six to eight hours have been used in patients with major bleeding, hemophilic bleeding, or after major trauma. After cardiac surgery, doses of 10 to 15 mg/kg as an intravenous bolus followed by 1 mg/kg per hour for 6 to 10 hours have been used. Tranexamic acid excretion is highly dependent on renal function; the interval between doses is substantively increased in patients with renal insufficiency.
•The dose of epsilon-aminocaproic acid depends on the urgency with which the bleeding needs to be reversed. A typical starting dose is 2 grams intravenously every six hours; as much as 1 gram intravenously every hour can be given. Epsilon-aminocaproic acid can also be administered orally at a dose of 3 grams three to four times per day.
●Desmopressin (DDAVP) can be used for impaired platelet function such as occurs in the setting of uremia or antiplatelet agents. Typical dosing is 0.3 mcg/kg given subcutaneously or intravenously (in 50 mL of normal saline over 15 to 30 minutes if given intravenously), which may improve platelet function for several hours. We usually do not give more than two doses because tachyphylaxis often develops after the second dose, and repeated administration of desmopressin can be complicated by hyponatremia.
High-quality data are lacking regarding the efficacy of these agents in the setting of DOAC-related bleeding. However, given their low risk of thrombosis, low cost, and widespread availability, these agents may be appropriate in patients with major or life-threatening DOAC-associated bleeding. A systematic review of non-surgical patients treated with tranexamic acid found no increase in the risk of thrombosis (venous or arterial) [57].
Additional details regarding the antifibrinolytic agents are presented separately. (See "Thrombotic and hemorrhagic disorders due to abnormal fibrinolysis", section on 'Tranexamic acid and aminocaproic acid'.)
Transfusions if needed — Transfusions are a component of supportive care for severe bleeding. (See "Initial management of moderate to severe hemorrhage in the adult trauma patient", section on 'Moderate to severe ongoing hemorrhage' and "Approach to shock in the adult trauma patient".)
●Red blood cells (RBCs) – RBC transfusion may be required, depending on the rate of bleeding and amount of blood loss. (See "Indications and hemoglobin thresholds for RBC transfusion in adults" and "Initial management of moderate to severe hemorrhage in the adult trauma patient".)
●Platelets – Platelet transfusion is not used to reverse the anticoagulant effect of DOACs in a patient with a normal platelet count; however, thrombocytopenic patients with bleeding should be treated for the underlying cause of thrombocytopenia and receive platelet transfusions if the thrombocytopenia is severe and bleeding is major or life-threatening. Platelet transfusions may also be used in patients with severe bleeding in the setting of antiplatelet medications such as aspirin. (See "Platelet transfusion: Indications, ordering, and associated risks".)
●Plasma products – Plasma products such as FFP or PF24 may be given as part of a massive transfusion protocol, to replace coagulation factors lost by bleeding and replacement of packed RBCs, which do not contain coagulation factors at replacement levels. In contrast, there is no evidence to support the use of FFP as a reversal strategy in DOAC-associated bleeding. (See 'Clotting factor products' above and "Massive blood transfusion" and "Use of blood products in the critically ill".)
MINOR BLEEDING — Minor bleeding (eg, epistaxis, bruising, slow gastrointestinal bleeding) can usually be managed conservatively using local hemostatic measures (eg, mechanical pressure). Decisions to temporarily discontinue the anticoagulant must balance bleeding and thrombotic risks for each patient on a case-by-case basis (table 2).
The use of antifibrinolytic agents may occasionally be appropriate in individuals with less serious bleeding if the patient has ongoing bleeding or other comorbidities that increase bleeding risk; however, these agents generally are not required for minor bleeding. We do not administer prothrombin complex concentrates (PCCs) for minor bleeding.
SURGERY/INVASIVE PROCEDURE — Given the short half-lives of the DOACs, it is often possible to delay a procedure long enough to allow most or all of the anticoagulant effect to dissipate spontaneously in patients with normal renal or hepatic function (depending on the specific drug and its means of elimination). However, if urgent or emergent surgery is required and there is insufficient time to allow the anticoagulant effect to dissipate, reversal strategies discussed herein may be appropriate. Decisions regarding the need for reversal are individualized based on the urgency and bleeding risk of the procedure. (See 'Anticoagulant reversal' above.)
Perioperative management of DOACs in the setting of elective surgery is discussed separately. (See "Perioperative management of patients receiving anticoagulants".)
Emergency surgical interventions should proceed regardless of anticoagulation status or the receipt of reversal agents or a reversal strategy. In such circumstances, strategies to ameliorate the anticoagulant effect can be implemented intraoperatively if required.
HEAVY MENSTRUAL BLEEDING — (See "Abnormal uterine bleeding in nonpregnant reproductive-age patients: Management", section on 'Patients on anticoagulant therapy'.)
RESUMPTION OF ANTICOAGULATION — In all patients, a plan to decide about restarting the anticoagulant should be put in place; resumption should be delayed until the risk of rebleeding is minimized and may be permanently deferred in selected patients with a very high risk of recurrent bleeding.
Most patients can have their anticoagulants restarted [58]. Several studies have suggested that failure to restart the anticoagulant is associated with increased risks of thrombosis and death [59-61].
As an example, a large database study from 2021 involving nearly 7000 adults >65 years of age who had a bleed while receiving an anticoagulant found that, compared with not resuming the anticoagulant, resuming the anticoagulant was associated with [62]:
●Lower risk of thrombosis (adjusted hazard ratio [HR] 0.60, 95% CI 0.50-0.72)
●Higher risk of rebleeding (HR 1.88, 95% CI 1.64-2.17)
●Lower mortality (adjusted HR 0.54, 95% CI 0.48-0.60)
Approximately 80 percent of patients were receiving anticoagulation for atrial fibrillation; approximately one-half were receiving a DOAC and one-half warfarin; and the most common sites of bleeding were gastrointestinal and intracranial bleeding [62]. Approximately 70 percent of patients resumed their anticoagulant, with a median time to resuming the anticoagulant of 46 days (interquartile range, 6 to 550 days). Individuals with a previous bleed (prior to the index bleed that led them to be enrolled into this study) were excluded.
The finding that restarting anticoagulation is associated with reduced thrombosis, reduced death, and increased recurrent bleeding has been known for some time and supports our practice of restarting anticoagulation after recovering from a bleeding episode, except in individuals with an ongoing risk for major or life-threatening recurrent hemorrhage or those who no longer require anticoagulation [58].
We are less likely to restart the anticoagulant in an individual with a low risk of first or recurrent thromboembolism or whose underlying lesion strongly predisposes to recurrent bleeding, such as a patient with an intracerebral bleed and "congophilic angiopathy" on magnetic resonance imaging (MRI) scan. Each decision regarding resumption of anticoagulation following a bleed is individualized based on the risks and benefits for the specific patient as well as the patient's values and preferences (table 5).
The choice of anticoagulant also depends on a number of patient and bleed-related features, as discussed separately. (See "Direct oral anticoagulants (DOACs) and parenteral direct-acting anticoagulants: Dosing and adverse effects", section on 'Comparison with heparin and warfarin'.)
If anticoagulants are to be restarted, this can generally be done sooner than was identified in the study mentioned above [62]. The optimal time for restarting anticoagulation after a gastrointestinal bleed appears to be 7 to 14 days, although details depend on the cause and site of bleeding and the risks of rebleeding and thrombosis. Additional information related to the site of bleeding or the underlying indication for anticoagulation are presented in topic reviews that address these considerations:
●Intracerebral hemorrhage – (See "Reversal of anticoagulation in intracranial hemorrhage", section on 'Resumption of anticoagulation'.)
●Gastrointestinal bleeding – (See "Management of anticoagulants in patients undergoing endoscopic procedures", section on 'Resuming anticoagulants after hemostasis' and "Overview of the treatment of bleeding peptic ulcers", section on 'Risk factors for persistent or recurrent bleeding'.)
●Atrial fibrillation – (See "Atrial fibrillation in adults: Use of oral anticoagulants".)
●Venous thromboembolism – (See "Selecting adult patients with lower extremity deep venous thrombosis and pulmonary embolism for indefinite anticoagulation" and "Venous thromboembolism: Initiation of anticoagulation".)
ANTIDOTES UNDER DEVELOPMENT — The following strategies are at various stages of development:
●A small molecule antidote (PER977; Ciraparantag) has been shown in preliminary studies to bind directly and specifically to direct thrombin inhibitors, factor Xa inhibitors, and heparins (including LMW heparin), and to reverse their anticoagulant properties [63]. Of the anticoagulants, edoxaban required the lowest dose for full reversal of effect. In a study of 80 healthy volunteers given a single therapeutic dose of edoxaban, PER977 normalized the whole blood clotting time within 10 minutes; in contrast, normalization of the clotting time took 12 to 15 hours in individuals given edoxaban followed by placebo [64]. There were no major adverse effects.
●FXaI16L is a mutant form of factor Xa that introduces an amino acid substitution (isoleucine for leucine) at position 16; this agent is also being explored as a possible universal bypassing agent for multiple anticoagulants [65]. Rather than binding to an anticoagulant, FXaI16L circulates in a zymogen-like (inactive) state, is resistant to active site inhibitors, and becomes activated when it encounters activated factor V (factor Va) on damaged cellular surfaces, restoring hemostasis selectively at the site of bleeding. In a mouse model, FXaI16L was able to improve clotting and reduce blood loss following rivaroxaban treatment; it could also reverse the effect of rivaroxaban or dabigatran in vitro in human plasma [65].
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Anticoagulation".)
SUMMARY AND RECOMMENDATIONS
●Risk of bleeding – Direct oral anticoagulants (DOACs) reversibly inhibit coagulation factors and have a shorter half-life than warfarin. Available agents include the direct thrombin inhibitor dabigatran (Pradaxa) and the direct factor Xa inhibitors rivaroxaban (Xarelto), apixaban (Eliquis), and edoxaban (Lixiana, Savaysa). The risk of major bleeding with DOACs is low and generally similar to or lower than other anticoagulants, but life-threatening hemorrhages have occurred. (See 'Definition of terms' above and "Risks and prevention of bleeding with oral anticoagulants".)
●Evaluation – We assess the severity of bleeding and the degree of anticoagulation through the patient history and physical examination. The degree of anticoagulation depends on the dose and the interval since the last dose; in some cases coagulation testing may be helpful in determining residual anticoagulant effect, but normal coagulation testing does not necessarily eliminate the need for aggressive interventions. Other limited laboratory testing includes a complete blood count (CBC) and tests of renal and hepatic function. (See 'Patient assessment' above.)
●Initial treatment – In most cases of DOAC-associated bleeding, including major bleeding, we discontinue the drug; transfuse blood products if necessary; address the hemorrhage anatomically (eg, surgery, endoscopy, local measures); and administer pro-hemostatic therapies such as antifibrinolytic agents (table 2). (See 'General strategy for anticoagulant reversal' above and 'Transfusions if needed' above and 'Antifibrinolytics and other pro-hemostatic therapies' above.)
Reversal agents – Agent-specific interventions may include the following:
•Dabigatran – For patients with major bleeding, including life-threatening bleeding (eg, intracranial, severe gastrointestinal), we suggest administration of a specific reversal agent (idarucizumab) (table 3) along with the antifibrinolytic agent (eg, tranexamic acid, epsilon-aminocaproic acid) (Grade 2C). We may also use an antifibrinolytic agent in selected patients with major bleeding that is not immediately life-threatening (eg, suspected overdose, comorbidities, worsening bleeding symptoms). We also suggest administration of oral activated charcoal if the last anticoagulant dose was within the previous two hours and the patient can tolerate enteral administration (ie, not vomiting, adequate mental status, or via gastric tube if tracheally intubated) (Grade 2C). Hemodialysis may be used in selected patients if the potential for significant drug removal is high. (See 'Dabigatran reversal' above.)
For patients at an imminent risk of death from bleeding for whom idarucizumab is not available, we suggest using an activated prothrombin complex concentrate (aPCC; eg, factor eight inhibitor bypassing agent [FEIBA] (table 4)) (Grade 2C). We would not give idarucizumab and an aPCC together, and we suggest not giving an aPCC for major bleeding that is not life-threatening (Grade 2C). This is because treatment with aPCCs carries a real and substantial prothrombotic risk, so we avoid these agents except in extreme clinical circumstances.
•Rivaroxaban, apixaban, edoxaban – For patients with major bleeding, including life-threatening bleeding (eg, intracranial, gastrointestinal), we suggest an antifibrinolytic agent (eg, tranexamic acid, epsilon-aminocaproic acid) (Grade 2C). We may also use an antifibrinolytic agent in selected patients with major bleeding that is not immediately life-threatening, (eg, suspected overdose, comorbidities, worsening bleeding symptoms). We also suggest administration of activated charcoal if the last dose of the anticoagulant was within the previous two hours and the patient can tolerate enteral administration (Grade 2C). The direct factor Xa inhibitors cannot be dialyzed. (See 'Factor Xa inhibitors' above.)
For patients at imminent risk of death from bleeding, in addition to treatment with an antifibrinolytic agent and drug removal with activated charcoal, we suggest administering andexanet alfa (table 3) or an unactivated 4-factor PCC (Grade 2C). Andexanet carries a Boxed Warning regarding arterial and venous thrombosis, treatment with PCCs carries a real prothrombotic risk, and the quality of the available evidence regarding the safety and efficacy of PCCs in factor Xa inhibitor-associated major bleeding is low. Therefore, we suggest not using these agents except in extreme clinical circumstances (Grade 2C).
●Minor bleeding – Minor bleeding can usually be managed conservatively using local hemostatic measures (ie, without anticoagulant reversal). Decisions to temporarily discontinue the anticoagulant must balance bleeding and thrombotic risks for each patient on a case-by-case basis. (See 'Minor bleeding' above.)
●Heavy menstrual bleeding – (See "Abnormal uterine bleeding in nonpregnant reproductive-age patients: Management", section on 'Patients on anticoagulant therapy'.)
●Surgery/invasive procedure – (See "Perioperative management of patients receiving anticoagulants", section on 'DOAC interruptions (overview)'.)
●Restarting the anticoagulant – Most patients can resume anticoagulation after a bleed, with accumulating evidence that restarting the anticoagulant results in improved outcomes with lower risks of thrombosis. These decisions are individualized based on the risks and benefits for the specific patient as well as the patient's values and preferences (table 5). (See 'Resumption of anticoagulation' above.)
●ICH-specific management – Prevention of bleeding and management of intracerebral hemorrhage (ICH) are discussed separately. (See "Risks and prevention of bleeding with oral anticoagulants" and "Reversal of anticoagulation in intracranial hemorrhage".)
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟
