Principal Developer: Mark Crowther
Secondary Developers: Don Houston, Agnes Lee, Mary-Frances Scully

Managing warfarin associated coagulopathy

Background

Warfarin is a highly effective antithrombotic which produces risk reductions for thrombosis in primary and secondary prevention of 60 to 95%. Warfarin is monitored using the International Normalized Ratio (INR); for most indications the recommended target INR is 2.0 to 3.0.

Despite careful monitoring, typically at least 30 - 40% of INR determinations are outside the desired therapeutic window. The recommendations contained in this guide apply to the management of those patients whose INR is "above" the usual therapeutic window. The management of patients who hemorrhage while receiving warfarin therapy is also discussed.

The degree of prolongation of the INR correlates with bleeding risk. Thus in general the intervention applied to a patient with a prolonged INR is contingent on two factors (a) the degree of prolongation of the INR and (b) the presence or absence of bleeding.

Treatment of patients with warfarin-associated coagulopathy has been the subject of a number of recent investigations and clinical trials.  The results of these studies are useful to guide clinical practice.

In all cases of excess warfarin effect consensus recommendations suggests that the etiology of the excess warfarin effect should be investigated and corrected where possible.  In many cases the prolonged INR will be attributable to either transient excess alcohol consumption, temporary introduction of medications such as antibiotics, significant changes in dietary vitamin K intake, or the initiation of new medications which will be used over an extended period of time.

Management of Non-bleeding patients

Minor deviations of the INR outside of the range of 2 to 3 in a generally stable patient do not need intervention; rechecking the INR is recommended.

INR 3.5 – 4.4

Consensus recommendations and sensible clinical practice suggest that in such patients the warfarin may be withheld for 24 hours then reinstituted at a lower dose with repeat INR monitoring. In many patients warfarin can be safely continued at the usual dose if the cause of the increased warfarin effect (for example, transient administration of an antibiotic) can be identified and addressed.  There is no evidence support for use of either vitamin K or plasma transfusion in such patients.

INR 4.5 to 10.0

To allow the INR to return towards the usual therapeutic window warfarin should be withheld for a period of at least 24 hours.  Warfarin can then be reintroduced at a lower dose if the cause of the elevated INR is either unapparent or cannot be corrected. If the cause can be identified and corrected, warfarin can be introduced at its previous dose.  Oral vitamin K will reduce the INR towards the therapeutic reference interval more quickly than can be achieved through simple warfarin withdrawal; in the absence of bleeding there is no evidence to support the use of either intravenous vitamin K or plasma in the form of fresh frozen plasma or prothrombin complex concentrates.

INR >10.0

Treatment of patients with markedly prolonged INR values is controversial and based largely on anecdotal experience.  Recent evidence suggests that the rate of major and/or life-threatening bleeding in such patients is low.  Consensus recommendations suggest that warfarin be withheld for a period of 48 to 72 hours and then reintroduced cautiously, particularly when the etiology of the elevation of the INR cannot be identified.  Oral vitamin K at a dose of 2.5 mg will rapidly reduce the INR towards the usual reference interval and is not known to cause either "overcorrection” or "rebound" thromboembolism.  It is intuitive that administration of vitamin K should reduce bleeding risk, though that has not been proven in clinical trials.  In the absence of bleeding there is no evidence to support the use of either intravenous vitamin K or transfusion of plasma products.  Although widely practiced, there is no evidence that hospital admission benefits such patients.  Most such patients can be managed safely through the outpatient department with the administration of oral vitamin K and attentive clinical follow-up.

Management of bleeding patients

Major or life threatening bleeding

Major or life-threatening bleeding in patients receiving warfarin should be regarded as a medical emergency.  The case fatality rate for patients admitted to the hospital with anticoagulant-associated bleeding is high; in a recent review Douketis and colleagues reported that 8% of patients with a major bleeding while receiving warfarin ultimately died as a consequence of that bleeding complication (1).

Therapy for patients with major bleeding has two principal aims: (a) rapidly correcting the coagulopathy to allow normal hemostasis and (b) administration of fluid and blood products to correct hypovolemia and anemia. Management of bleeding patients should be consistent irrespective of the INR. Even when the INR is in the therapeutic range, warfarin will increase bleeding.

Major hemorrhage requires interruption of warfarin therapy, medical stabilization and administration of lifesaving therapies, transfusional therapy to correct the warfarin-associated coagulopathy and any associated dilutional coagulopathy, and the administration of intravenous vitamin K at doses of 2.5 to 5 mg.

Although recommendations have been made for specific volumes of plasma to correct warfarin-associated coagulopathy, such calculations have not been validated in prospective clinical trials. A more circumspect clinical practice is to administer two to four units of fresh frozen plasma (500 to 1000 ml) and then reassess the INR.

Prothrombin complex concentrates probably more rapidly correct the INR; however their impact on clinical outcomes has not been determined. PCC may have particular value in patients with cardiac or renal disease who cannot tolerate large volumes of plasma, or in patients with immediately life threatening bleeding where the speed with which PCC can be administered is a major advantage. They may be also of particularly importance where immediate INR correction may be beneficial, such as in patients with immediately life-threatening hemorrhage such as intracerebral bleeding.

Intravenous vitamin K has been associated with anaphylactic reactions attributable, in most cases, to the diluents used to make the product miscible.  The risk of anaphylaxis is much lower with contemporary formulations using propylene glycol and polyethylene glycol than with historical castor oil-based formulations.

Minor bleeding

Patients with minor bleeding (such as epistaxis or menorrhagia) can usually be managed with local measures, with or without temporary warfarin interruption.  Persistent minor bleeding may require more prolonged interruption of warfarin therapy and/or interventions such as cauterization. Lowering the target INR does not, in general, reduce bleeding but does reduce the antithrombotic efficacy of warfarin.

Acquisition and administration of oral Vitamin K

Vitamin K is available in Canada only in ampoules at a concentration of 10mg/ml or 1 mg/ml. Small doses can be drawn up with a graduated 1ml syringe and diluted with water or juice to drink.

Reinstituting warfarin after a major bleeding episode

When major bleeding occurs, every effort should be made to find a treat a reversible cause. After recovering from a major bleeding episode, reinstitution of warfarin may be considered if the cause of bleeding is reversible and correctable. When considering reintroduction, the clinician should consider both the risk of recurrent bleeding and the risk of thrombosis borne by the patient.

Advice from appropriate specialists should be sought.

Although not based on evidence, a commonly employed practice is to gradually re-anticoagulate patients with low doses of heparin, and, provided there is no recurrent bleeding, gradually transition to warfarin.

Reference List

  (1)   Linkins LA, Choi PT, Douketis JD. Clinical impact of bleeding in patients taking oral anticoagulant therapy for venous thromboembolism: a meta-analysis.[see comment]. Ann Intern Med 2003 Dec 2;139(11):893-900.