Library › Pharmacology › Warfarin

Warfarin: dosing, INR management, and reversal

Warfarin from the vitamin K cycle and pharmacokinetics through indications, INR targets, initiation and maintenance dosing, interactions, pharmacogenomics, supratherapeutic INR, adverse effects, and special populations.

Mechanism of action: the vitamin K cycle and VKORC1

The coagulation factors II (prothrombin), VII, IX, and X and the anticoagulant proteins C and S are synthesized in the liver as precursors that are inactive until glutamic acid residues near their amino termini are converted to γ-carboxyglutamic acid (Gla). The γ-glutamyl carboxylase that performs this step requires reduced vitamin K (vitamin K hydroquinone) as a cofactor. With each carboxylation, the hydroquinone is oxidized to vitamin K 2,3-epoxide. The epoxide is recycled to the quinone and then to the hydroquinone by vitamin K epoxide reductase, whose catalytic subunit is encoded by VKORC1. The Gla residues bind calcium and allow the factors to assemble on phospholipid membranes; undercarboxylated factors circulate but have little procoagulant activity.

Warfarin inhibits vitamin K epoxide reductase complex subunit 1. Recycling of the epoxide fails, the hepatic pool of reduced vitamin K falls, and newly synthesized vitamin K-dependent proteins are released in partly carboxylated, functionally defective form. Warfarin has no effect on factors already in the circulation. Its anticoagulant effect therefore develops only as existing functional factors are cleared, at rates set by their half-lives.

Dietary vitamin K (phylloquinone) can be reduced to the hydroquinone by a second, warfarin-insensitive reductase. This pathway explains two clinical observations. Large or variable intake of vitamin K lowers the response to a given warfarin dose, and pharmacologic vitamin K reverses warfarin even while the drug is still present. High intravenous doses of vitamin K saturate the liver and can make the patient resistant to warfarin for a week or more after reversal.

Warfarin is a coumarin derivative, one of several vitamin K antagonists. Acenocoumarol and phenprocoumon are used in Europe and elsewhere and differ mainly in half-life. Warfarin is the only vitamin K antagonist marketed in the United States, where the original branded product has been discontinued and generic warfarin and Jantoven remain.

Time course of anticoagulation: factor half-lives and early protein C depletion

The prothrombin time begins to rise within about 24 hours of the first dose, but the peak anticoagulant effect is delayed 72 to 96 hours, and a single dose acts for 2 to 5 days. The early rise in INR reflects the fall in factor VII, which has the shortest half-life of the procoagulant factors. The antithrombotic effect depends mainly on the reduction of prothrombin and factor X, which takes several days. An INR in the therapeutic range on day 2 or 3 therefore reflects factor VII depletion and does not indicate protection against thrombus extension.

Half-lives from the US prescribing information for warfarin, section 12.2. Factor VII and protein C half-lives are short, so a large initial dose lowers both quickly without a faster fall in prothrombin.

Protein C falls in parallel with factor VII and well before prothrombin and factor X. For the first several days, the balance between procoagulant and anticoagulant proteins can shift toward thrombosis. This transient hypercoagulable state is the mechanism of warfarin-induced skin necrosis, most often in patients with hereditary or acquired protein C or protein S deficiency, and of venous limb gangrene when warfarin is started during acute heparin-induced thrombocytopenia. It is also one of two reasons that acute venous thromboembolism is treated with a parenteral anticoagulant until warfarin has been given for at least 5 days. The other reason is the slow fall of prothrombin.

The same kinetics apply on stopping. After the last dose, the INR falls over several days as the liver resynthesizes functional factors. Most patients with an INR in the therapeutic range reach an INR below 1.5 about 5 days after the last dose. Restarting warfarin after a procedure produces a therapeutic INR only after several days, which is the basis for heparin bridging in the small group of patients at very high thrombotic risk.

Pharmacokinetics: absorption, protein binding, enantiomers, and metabolism

Warfarin is essentially completely absorbed after oral administration, with peak plasma concentrations within about 4 hours. It has a small volume of distribution of about 0.14 L/kg because about 99% of the drug is bound to plasma proteins, chiefly albumin. Only the unbound fraction is active and available for metabolism. Displacement from albumin by another drug raises the free fraction only transiently, because clearance of free drug rises in parallel. Interactions mediated by protein binding are therefore of minor importance compared with those mediated by metabolism.

Commercial warfarin is a racemic mixture of S- and R-enantiomers. S-warfarin has 2 to 5 times the anticoagulant activity of R-warfarin and is cleared more rapidly. S-warfarin is metabolized chiefly by cytochrome P450 (CYP) 2C9 to inactive hydroxylated metabolites. R-warfarin is metabolized by CYP1A2 and CYP3A4, with minor contributions from other isoenzymes and from reductases. Because the S-enantiomer carries most of the effect, inhibitors of CYP2C9 produce large changes in the INR, while drugs that act only on the R-enantiomer, such as cimetidine and omeprazole, produce modest changes. Amiodarone inhibits both pathways.

Values from the US prescribing information for warfarin (sections 8 and 12.3) and the American College of Chest Physicians (ACCP) pharmacology chapters.

Maintenance requirements vary more than 10-fold between patients, from under 1 mg to more than 15 mg daily. Age, body size, genotype, diet, liver function, thyroid function, and concomitant drugs explain much of this variation. Older patients need lower doses, and the prescribing information notes that patients 60 years or older show a greater than expected INR response. Asian patients, who more often carry the VKORC1 variant associated with warfarin sensitivity, need lower initiation and maintenance doses; in a study of Chinese outpatients, the mean requirement was 3.3 mg daily. Hereditary resistance from VKORC1 coding mutations is rare. Most patients who appear resistant are not taking the drug as prescribed or have a high vitamin K intake.

Measuring warfarin effect: prothrombin time, ISI, and the INR

The prothrombin time is the clotting time of citrated plasma after addition of calcium and thromboplastin, a reagent containing tissue factor and phospholipid. It is prolonged by reduction of factors II, VII, and X and also by reduction of fibrinogen and factor V, which warfarin does not affect. Thromboplastins differ in their sensitivity to the reduction of these factors, so the same plasma gives different prothrombin times in different laboratories.

The INR standardizes the result. Each thromboplastin is assigned an international sensitivity index (ISI) by calibration against an international reference reagent. The INR is calculated as INR = (patient prothrombin time ÷ mean normal prothrombin time)^ISI. A highly sensitive reagent has an ISI near 1.0. The mean normal prothrombin time is determined locally for each reagent and instrument combination, and an incorrect ISI or mean normal value is a recognized source of INR error.

The INR was validated in patients on stable warfarin therapy. Its interpretation is less certain in several settings:

• Early in therapy. The INR during the first days reflects factor VII more than prothrombin, and the INR system is less accurate in this phase.
• Liver disease. The INR is not standardized for the coagulopathy of cirrhosis and does not measure bleeding risk in that setting. Baseline elevation complicates warfarin dosing.
• Lupus anticoagulant. Antiphospholipid antibodies can prolong phospholipid-dependent clotting times. Most patients with a lupus anticoagulant have a normal baseline prothrombin time, and a target INR of 2.0 to 3.0 is used. In the minority with a prolonged baseline prothrombin time, a laboratory INR with a reagent insensitive to the lupus anticoagulant or a chromogenic factor X assay is used. The same applies when point-of-care INR results disagree with laboratory results. A chromogenic factor X activity of about 20% to 40% of normal corresponds to an INR of 2.0 to 3.0, and each laboratory validates its own range.
• Other anticoagulants. Unfractionated heparin raises the INR modestly. When heparin and warfarin overlap, the prescribing information specifies INR sampling at least 5 hours after an intravenous bolus, 4 hours after stopping an infusion, or 24 hours after a subcutaneous dose. Argatroban raises the INR substantially, which complicates the transition from argatroban to warfarin. Rivaroxaban, edoxaban, and dabigatran raise the INR, and apixaban raises it less, which matters when switching from a DOAC to warfarin.
• High values. INR values above about 4.5 are less reliable than values near the therapeutic range, and decisions at high INR values are approximate.

Point-of-care devices measure the INR in capillary whole blood. They allow testing at home and in clinics with immediate dose decisions, and their results agree with laboratory values within the therapeutic range. Agreement is weaker at high INR values, in the presence of a lupus anticoagulant, and with hematocrit outside the device range, so high or unexpected point-of-care results are confirmed with a venous sample.

Indications where warfarin remains the anticoagulant of choice

Warfarin is approved for prophylaxis and treatment of venous thrombosis and pulmonary embolism, for thromboembolic complications of atrial fibrillation and cardiac valve replacement, and to reduce death, recurrent myocardial infarction, and thromboembolism after myocardial infarction. Four randomized programs in atrial fibrillation and four in venous thromboembolism established DOACs as at least as effective as warfarin with less intracranial hemorrhage. Both the 2023 American College of Cardiology (ACC)/American Heart Association (AHA)/ACCP/Heart Rhythm Society (HRS) atrial fibrillation guideline and the 2021 CHEST venous thromboembolism guideline recommend a DOAC over warfarin for eligible patients. The direct oral anticoagulants page covers those trials and DOAC dosing.

Warfarin remains the required or preferred anticoagulant in the settings below. The first three rest on randomized trials of a DOAC that showed harm or failed to show adequate protection. The others rest on smaller trials, observational data, or device protocols.

Trial results from the primary publications. A DOAC is recommended over a vitamin K antagonist for atrial fibrillation with other native valve disease, after mitral repair, and with a bioprosthesis implanted more than 3 months earlier.

Duration of therapy. The duration depends on the indication, not on the drug. Mechanical valves, atrial fibrillation with moderate or severe rheumatic mitral stenosis, and thrombotic antiphospholipid syndrome require indefinite anticoagulation. Venous thromboembolism provoked by a major transient risk factor is treated for 3 months. Unprovoked thrombosis and thrombosis with a persistent risk factor are considered for extended therapy after the first 3 months, with periodic reassessment of bleeding risk (CHEST 2021). The decision is covered on the pulmonary embolism page. Left ventricular thrombus is treated for at least 3 months and until resolution on repeat imaging.

The common feature of the mechanical valve, moderate or severe rheumatic mitral stenosis, and antiphospholipid syndrome is thrombus formation that a single-target inhibitor does not suppress at tolerable doses. A mechanical prosthesis activates coagulation through contact activation on its artificial surfaces and through high shear at hinge points and the sewing ring. Moderate or severe rheumatic mitral stenosis produces marked left atrial stasis with a fibrin-rich thrombus burden. Antiphospholipid antibodies activate endothelium, monocytes, platelets, and complement, with tissue factor expression that drives coagulation from several entry points. Warfarin lowers four procoagulant factors and reduces thrombin generation at more than one level of the cascade. Each indication is established by history and testing. A history of valve surgery identifies a mechanical prosthesis, echocardiography defines mitral valve morphology and area in patients from regions where rheumatic fever is endemic, and antiphospholipid antibody testing follows unprovoked or arterial thrombosis.

INR targets by indication

The target INR of 2.5 (range 2.0 to 3.0) rests on randomized comparisons of intensity and on observational analyses of event rates by INR. In atrial fibrillation, an INR below 2.0 at the time of ischemic stroke is associated with more severe strokes (odds ratio 1.9) and higher 30-day mortality (hazard ratio 3.4) than an INR of 2.0 or more. The risk of intracranial hemorrhage rises steeply above an INR of about 4.0. Higher-intensity regimens increase bleeding without reducing thrombosis in most indications. The 2012 ACCP guideline recommended a target of 2.5 (range 2.0 to 3.0) for patients treated with a vitamin K antagonist over lower or higher ranges (Grade 1B).

The 2020 ACC/AHA valve guideline specifies a single target with an acceptable range of 0.5 INR units on each side, which reduces time near the boundaries of the range. Mechanical valve anticoagulation, valve thrombosis, and bridging are covered on the prosthetic heart valves page. TTR, time in therapeutic range.

The low INR range for the On-X aortic valve in the table applies to patients without thromboembolic risk factors. The evidence comes from the high-risk arm of the Prospective Randomized On-X Valve Anticoagulation Clinical Trial (PROACT), which enrolled 375 patients with an On-X aortic valve and at least one thromboembolic risk factor. After standard warfarin for 3 months, they were randomized to an INR of 1.5 to 2.0 or 2.0 to 3.0, all with aspirin 81 mg daily. In this arm, the lower target reduced major and minor bleeding without an increase in stroke, transient ischemic attack, or death. The low-risk arm, in patients without risk factors, did not test a lower INR. It compared aspirin plus clopidogrel with standard warfarin plus aspirin and was stopped early for excess cerebral thromboembolism with dual antiplatelet therapy. The labeling that the US Food and Drug Administration (FDA) approved for the On-X aortic valve permits an INR of 1.5 to 2.0 with aspirin 81 mg daily from 3 months after implantation. The 2020 ACC/AHA valve guideline gives this lower target a Class 2b, B-R recommendation and limits it to patients without thromboembolic risk factors, although the randomized comparison of INR targets was made in patients with risk factors. The strategy applies only to the On-X aortic prosthesis and is not extrapolated to other valves or to the mitral position, where the target remains 3.0 (2.5 to 3.5). Aspirin is not added to warfarin routinely for every mechanical valve; it is used when there is an atherosclerotic indication or after thromboembolism despite a therapeutic INR.

Contraindications and cautions

The US prescribing information lists warfarin as contraindicated in:

• pregnancy, except in women with mechanical heart valves at high thromboembolic risk
• hemorrhagic tendencies or blood dyscrasias
• recent or planned surgery of the central nervous system or eye
• traumatic surgery with large open surfaces
• active ulceration or overt bleeding of the gastrointestinal, genitourinary, or respiratory tract
• central nervous system hemorrhage
• cerebral aneurysm
• aortic dissection
• pericarditis and pericardial effusion
• bacterial endocarditis
• threatened abortion
• eclampsia and preeclampsia
• unsupervised patients with conditions associated with a high likelihood of nonadherence
• spinal puncture and other procedures with potential for uncontrollable bleeding
• major regional or lumbar block anesthesia
• malignant hypertension
• hypersensitivity to warfarin

Many of these are relative in practice. A patient with a mechanical valve and infective endocarditis, for example, usually continues anticoagulation unless there is a cerebral embolic or hemorrhagic complication, with the decision made by the endocarditis team.

Added risk accompanies moderate to severe hepatic impairment, disturbances of intestinal flora from infection or antibiotics, indwelling catheters, moderate to severe hypertension, and known protein C or protein S deficiency. Concomitant heparin for 5 to 7 days during initiation may reduce the risk of skin necrosis in patients with protein C or S deficiency. Warfarin is not used as initial therapy in acute heparin-induced thrombocytopenia. A history of falls, advanced age, and frailty are not contraindications in themselves, because the absolute stroke reduction in older patients with atrial fibrillation usually exceeds the added bleeding risk.

Starting warfarin: dose selection and parenteral overlap

The prescribing information recommends an initial dose of 2 to 5 mg daily when genotype is unknown, with typical maintenance doses of 2 to 10 mg daily, and it advises against routine loading doses. Guidelines allow a higher start in selected patients. The 2008 ACCP guideline recommended initiation with 5 to 10 mg for the first 1 or 2 days for most patients (Grade 1B). It also recommended a starting dose of 5 mg or less in older adults and in patients who are debilitated or malnourished, have heart failure or liver disease, have had recent major surgery, or take drugs that increase sensitivity such as amiodarone (Grade 1C). The 2012 ACCP guideline suggested 10 mg daily for the first 2 days in outpatients well enough to be treated at home, followed by INR-based dosing (Grade 2C).

Trials of starting doses reached different conclusions in different populations. In hospitalized patients, two small trials compared 5-mg and 10-mg starts. In the first, 44% of patients given 10 mg and 8% given 5 mg had an INR above 2.0 at 36 hours, reflecting rapid factor VII depletion, and 36% versus none had an INR above 3.0 at 60 hours. In the second, a stable INR of 2.0 to 3.0 by day 4 or 5 without overshoot was reached by 24% with 10 mg and 66% with 5 mg. In 201 outpatients with acute venous thromboembolism treated with low-molecular-weight heparin, a 10-mg nomogram brought 83% of patients to a therapeutic INR by day 5, compared with 46% with a 5-mg nomogram. It also shortened the overlap with low-molecular-weight heparin. The number of INR values above 5.0 did not differ between the groups, and recurrent venous thromboembolism at 90 days occurred in 3 patients with 10 mg and none with 5 mg, a difference that was not significant. A 10-mg start is used only in young, otherwise healthy outpatients without interacting drugs.

Parenteral overlap. In acute venous thromboembolism, warfarin is started on the first day of parenteral anticoagulation, and the parenteral agent continues for at least 5 days and until the INR has been 2.0 or more for at least 24 hours. Five days is the minimum even if the INR reaches range sooner. The prescribing information describes overlap of 4 to 5 days until the desired INR response. Patients with protein C deficiency, a history of skin necrosis, or heparin-induced thrombocytopenia need overlap under specialist supervision.

Atrial fibrillation without an acute thrombus does not require heparin during warfarin initiation. A lower starting dose (often 5 mg or less) and weekly INR measurement during the first month avoid early over-anticoagulation. When warfarin follows mechanical valve surgery, practice follows the surgical program. Intravenous heparin or low-molecular-weight heparin until the INR is therapeutic is common for mitral and high-risk aortic prostheses. Patients with a bileaflet aortic valve and no risk factors are often managed without parenteral overlap.

Baseline assessment. A baseline INR, complete blood count, and liver and kidney function are measured before the first dose. The baseline INR identifies liver disease, vitamin K deficiency, and lupus anticoagulants that prolong the prothrombin time. A medication and supplement reconciliation identifies interacting drugs before the dose is chosen.

Warfarin initiation nomograms

Validated nomograms reduce variation in dosing during initiation. Three are used most often. The 5-mg nomogram was developed in hospitalized patients with venous thromboembolism receiving heparin and adjusts the dose daily from day 3 to day 6. The 10-mg nomogram was tested in outpatients with venous thromboembolism receiving low-molecular-weight heparin and gives 10 mg on days 1 and 2, with INR measurements only on days 3 and 5. The day-5 nomogram estimates the maintenance dose in outpatients with nonrheumatic atrial fibrillation who take 5 mg on days 1 through 4 without heparin.

The 5-mg nomogram gives 5 mg on days 1 and 2; the doses for days 3 to 6 are shown in the table below. The day-5 estimate interpolates between rows of the published table for the 2001 outpatient scheme. Neither nomogram applies to patients taking amiodarone or other strong potentiators, or to patients with liver disease or a known sensitive genotype.

The 5-mg initiation nomogram, developed in inpatients with venous thromboembolism receiving heparin (1999 loading-dose trial, as adapted in the 2003 outpatient 10-mg versus 5-mg trial).

Day-5 maintenance estimate after 5 mg daily on days 1 to 4 in outpatients with nonrheumatic atrial fibrillation not receiving heparin (2001 outpatient scheme). In the prospective validation group of 23 patients, the mean difference between predicted and final dose was 1.6 mg per week.

A patient whose INR is above 3.0 on day 3 is unusually sensitive, from genotype, liver dysfunction, poor intake, or an interacting drug, and needs a much lower maintenance dose than the nomogram predicts. A patient whose INR has not moved by day 5 on 5 mg daily may not be taking the drug, may have a high vitamin K intake, or may need a higher dose. Pharmacogenetic algorithms, described below, are an alternative where genotype is available before the first dose.

Maintenance dosing and weekly dose adjustment

After initiation, dosing is managed in total weekly milligrams. Small adjustments are not possible with daily tablets alone, and a weekly schedule allows changes of 5% to 15% using combinations of tablet strengths on different days. US tablets are color-coded by strength (1, 2, 2.5, 3, 4, 5, 6, 7.5, and 10 mg), and a single tablet strength with a simple weekly pattern reduces dosing errors.

The principles of maintenance adjustment are consistent across validated algorithms:

1. INR in range. No change.
2. Single INR slightly out of range in a previously stable patient. The 2012 ACCP guideline suggested continuing the current dose and retesting within 1 to 2 weeks when a stable patient has one INR no more than 0.5 above or below the range (Grade 2C). An explanation (a missed or extra dose, a change in diet, alcohol, acute illness, a new drug) is sought before the dose is changed.
3. INR persistently or substantially out of range. The weekly dose is changed by about 10% to 15%. In RE-LY, the warfarin algorithm specified no change for in-range values and weekly dose changes of 10% to 15% for out-of-range values. Each 10% increase in a center's proportion of algorithm-consistent dose decisions predicted a 6.12 percentage-point increase in TTR. The degree of algorithm consistency explained 87% of the between-center variation in TTR.
4. Moderately high INR without bleeding. One dose is withheld and the weekly dose is reduced. Management of higher values is described in the section on supratherapeutic INR.
5. Low INR. The weekly dose is increased by about 10% to 15% after a cause such as missed doses or increased vitamin K intake is excluded. Heparin bridging for a single low INR in a stable patient is not suggested (ACCP 2012, Grade 2C), except in selected patients with a mechanical mitral valve or recent thromboembolism, in whom the decision is individualized.

The 2012 ACCP guideline suggested validated decision support (paper nomograms or computer programs) over unstructured dosing during maintenance (Grade 2C). Because a dose change takes about a week to reach a new steady state, the INR is rechecked after any change. Further changes are not made on the basis of an INR drawn 2 or 3 days after the previous adjustment unless it is high. Routine low-dose vitamin K supplementation to stabilize the INR is not suggested (ACCP 2012, Grade 2C).

INR monitoring frequency and recall intervals

Extended intervals apply to patients with consistently stable INRs on an unchanged dose, which some programs define as at least 3 months of results in range.

Extending the recall interval reduces the burden of testing without loss of control in stable patients. The ASH guideline judged the evidence very uncertain and made a conditional recommendation. Patients with a mechanical mitral valve, a history of labile INR, or a high-risk drug regimen are usually kept at shorter intervals.

Time in therapeutic range and the quality of anticoagulation

Time in therapeutic range (TTR) is the standard measure of warfarin control. The Rosendaal method assumes that the INR changes linearly between consecutive measurements and assigns an interpolated INR to each day. The percentage of days with an interpolated INR in range is the TTR. Simpler measures, such as the fraction of INR values in range, depend on how often the INR is measured.

TTR predicts outcomes. In atrial fibrillation trials and cohorts, lower TTR is associated with more stroke, more bleeding, and higher mortality. The 2023 ACC/AHA atrial fibrillation guideline describes a TTR of 70% or more as the goal when warfarin is used. A TTR below 60% defines labile INR in the HAS-BLED score (hypertension, abnormal kidney or liver function, stroke, bleeding, labile INR, elderly age, drugs or alcohol). In the DOAC trials, TTR in the warfarin arms ranged from a mean of 55% in ROCKET AF to a median of 68.4% in ENGAGE AF-TIMI 48. Patients with persistently poor control who have no warfarin-specific indication are switched to a DOAC; patients who must stay on warfarin receive more intensive management. The 2024 European Society of Cardiology (ESC) atrial fibrillation guideline likewise sets a goal of keeping the INR in the 2.0 to 3.0 range more than 70% of the time, and it advises switching from a vitamin K antagonist to a DOAC in eligible patients with poor INR control or a high risk of intracranial hemorrhage. The 2023 US guideline recommends DOACs over warfarin for eligible patients without making TTR the trigger.

Derived in the AFFIRM warfarin population and validated externally (c-index 0.72 internally and 0.70 externally). The race item reflects the derivation data and is a marker of unmeasured factors such as genotype, access to care, and diet; it does not by itself determine treatment.

The SAMe-TT₂R₂ score estimates the probability of good INR control before warfarin is started. In the derivation study, scores of 0 or 1 identified patients likely to do well on a vitamin K antagonist, and scores of 2 or more identified patients who needed additional measures to reach an acceptable TTR. The score is used when a choice between warfarin and a DOAC remains open, and to plan the intensity of management when warfarin is required.

Anticoagulation clinics, self-testing, and self-management

Warfarin outcomes depend on the system that manages it. Systematic management incorporates patient education, scheduled INR testing, tracking of results and missed tests, timely dose decisions, and clear communication of each decision to the patient. Specialized anticoagulation management services, usually led by pharmacists or nurses working under protocol, are suggested over usual care by the patient's own clinician (ASH 2018, conditional). Supplementary education beyond basic instruction is also suggested.

Self-testing and self-management. Patient self-testing means the patient measures the INR at home with a point-of-care device and reports it to the managing service. Patient self-management means the patient also adjusts the dose using an algorithm. The ASH 2018 guideline recommended self-management over other approaches for patients able to perform it (strong recommendation, low certainty) and suggested self-testing over clinic or laboratory testing for others. The 2012 ACCP guideline suggested self-management for motivated patients who demonstrate competence (Grade 2B). The 2025 ESC/European Association for Cardio-Thoracic Surgery (EACTS) valve guideline recommends INR self-monitoring or self-management in selected trained patients with mechanical valves.

Self-management requires training, demonstrated competence with the device and the dosing algorithm, and a managing service that reviews results.

Self-management benefits younger, motivated patients with long-term indications, especially those with mechanical valves. Patients unable to perform testing, or with cognitive impairment, are managed by an anticoagulation service. Quality metrics for a warfarin program include TTR by patient and by program, the proportion of patients with an INR above 4 or 5, the interval between an out-of-range result and the next test, rates of major bleeding and thromboembolism, the proportion of patients overdue for testing, and documentation of indication, target, and duration.

Drug interactions: mechanisms and management

Drug interactions with warfarin act through four mechanisms: inhibition or induction of CYP2C9 (S-warfarin) and of CYP1A2 and CYP3A4 (R-warfarin); altered vitamin K availability or factor catabolism; additive effects on platelet function or mucosal integrity; and effects on absorption. A systematic review of reported interactions found that 72% described potentiation and that 84% of reports were of poor quality, most of them single cases. Reports were consistent for azole antifungals, macrolides, fluoroquinolones, nonsteroidal anti-inflammatory drugs (NSAIDs) including selective cyclooxygenase-2 inhibitors, selective serotonin reuptake inhibitors, omeprazole, lipid-lowering drugs, amiodarone, and fluorouracil.

Compiled from the US prescribing information (Tables 2 and 3), the ACCP 2008 interaction table graded by level of evidence, a systematic review of warfarin interactions, and a long-term cohort study of amiodarone. Antibiotics have inconsistent pharmacokinetic effects in studies, and the infection itself raises the INR, so INR monitoring is required with any antibiotic.

Amiodarone. Amiodarone and its metabolite desethylamiodarone inhibit CYP2C9 and also CYP1A2 and CYP3A4, so S- and R-warfarin clearance both fall. Because of amiodarone's very long half-life, the interaction develops over weeks and persists for weeks to months after amiodarone is stopped. In a cohort of 43 patients followed for at least a year, the interaction peaked at about 7 weeks with a mean maximum warfarin dose reduction of 44%. The required reduction correlated with the amiodarone maintenance dose: about 40% for 400 mg, 35% for 300 mg, 30% for 200 mg, and 25% for 100 mg daily. A preemptive dose reduction when amiodarone is started, with weekly INR monitoring through the first 2 months, prevents the high INR values that otherwise appear several weeks after the start.

Antiplatelet therapy. Aspirin increases bleeding when added to warfarin even at 75 to 100 mg daily, and the combination is used only when a separate indication exists, such as recent acute coronary syndrome, a coronary stent, or selected mechanical valve situations. The 2012 ACCP guideline suggested avoiding concomitant antiplatelet agents except when benefit is known or highly likely to exceed harm (Grade 2C). Combination therapy after percutaneous coronary intervention favors a DOAC with clopidogrel over warfarin-based regimens (2023 AF guideline, Class 1). Durations and trials are covered on the antithrombotic therapy in atrial fibrillation with PCI page.

Diet, vitamin K intake, alcohol, and supplements

Dietary vitamin K. Phylloquinone (vitamin K₁) comes mainly from green leafy vegetables (kale, spinach, collard and turnip greens, Brussels sprouts, broccoli), some vegetable oils, and green tea. Higher intake lowers the INR, and lower intake raises it. Patients are advised to keep a consistent intake, not to avoid these foods. Avoidance makes the INR more sensitive to small fluctuations, and in a case-control study of outpatients with an INR above 6.0, higher vitamin K intake was associated with lower risk (odds ratio 0.7 per unit of intake). Enteral feeds and some nutritional supplements contain vitamin K; changes in feeding regimens in hospitalized patients produce large INR swings. The 2023 AF guideline cited a trial in which weekly dietary vitamin K counseling increased the proportion of patients in range at 90 days from 58% to 74%. Reduced intake from anorexia, fasting, or acute illness is a common cause of a high INR.

Alcohol. Chronic heavy alcohol use induces hepatic enzymes and can lower the INR, while binge drinking, malnutrition, and alcohol-related liver disease raise it and increase bleeding risk. Light to moderate habitual consumption (1 drink every other day to 2 drinks daily) was associated with a lower risk of an INR above 6.0 in the outpatient case-control study, and ingestion of moderate amounts of wine had little effect on the prothrombin time in volunteers. Consistency of intake matters more than abstinence for INR stability, and alcohol excess is a modifiable bleeding risk factor in HAS-BLED.

Botanicals and supplements. Patients often do not report supplements, and about 30% of patients taking warfarin use herbal or natural product supplements regularly. The prescribing information lists garlic and ginkgo as having antiplatelet or anticoagulant effects. It lists St. John's wort, ginseng, and coenzyme Q10 as reducing the warfarin effect, and echinacea, grapefruit juice, ginkgo, goldenseal, and St. John's wort as possible CYP interactions. Fish oil at high doses has antiplatelet effects, and dong quai and danshen potentiate warfarin. More frequent INR monitoring is required whenever a botanical product is started or stopped, and medication reconciliation asks specifically about supplements.

Intercurrent illness and other causes of INR instability

Changes in the patient's condition also alter warfarin response. In the case-control study of outpatients with an INR above 6.0, independent risk factors were acetaminophen (odds ratio about 10 for 9.1 g per week or more), a new potentiating drug (8.5), advanced malignancy (16.4), recent diarrheal illness (3.5), decreased oral intake (3.6), and intake of more warfarin than prescribed (8.1).

From the ACCP 2008 pharmacology chapter, the US prescribing information, and the outpatient case-control study of excessive anticoagulation.

Hospitalization combines several of these influences: acute illness, poor intake, antibiotics, procedures, and changes in other drugs. INR values are measured daily or every other day in hospitalized patients on warfarin, and the discharge dose is not assumed to equal the previous outpatient dose.

Pharmacogenomics: CYP2C9, VKORC1, CYP4F2, and ancestry

Common genetic variants explain a large share of the variation in warfarin dose. In patients of European ancestry, VKORC1 genotype explains up to 30%, CYP2C9 up to 18%, and CYP4F2 up to 11% of the variance in stable dose. With clinical factors (age, body size, interacting drugs), algorithms explain about half the variance.

• VKORC1. The −1639G>A promoter variant (rs9923231) lowers expression of the drug target and increases warfarin sensitivity. Its frequency varies widely by ancestry and is highest in East Asian populations, which explains much of their lower dose requirement.
• CYP2C9. The 2 (R144C) and 3 (I359L) alleles reduce S-warfarin clearance, 3 more than 2. Carriers need lower doses, take longer to reach a stable INR, and have more over-anticoagulation and bleeding during initiation. The prescribing information notes that patients with 1/3, 2/2, 2/3, or 3/3 may need more than 2 to 4 weeks to reach the maximum INR effect of a given dose. The 5, 6, 8, and 11 alleles are found mainly in people of African ancestry and also reduce clearance.
• CYP4F2. The *3 variant (V433M, rs2108622) reduces metabolism of vitamin K₁ and slightly increases the dose requirement.
• rs12777823. A variant in the CYP2C cluster is associated with a lower dose requirement in African Americans.

Expected maintenance daily doses by genotype from the US prescribing information (Table 1), derived from several published studies. The table does not include CYP2C9*5, *6, *8, or *11, CYP4F2, or clinical factors, which algorithms add.

The Clinical Pharmacogenetics Implementation Consortium (CPIC) 2017 guideline addresses how to use genotype when it is available. For patients not of African ancestry, the dose is calculated with a published pharmacogenetic algorithm (the International Warfarin Pharmacogenetics Consortium algorithm or the Gage algorithm) using VKORC1 −1639G>A and CYP2C92 and 3 (strong recommendation). Carriers of CYP2C95, 6, 8, or 11 receive a further reduction of 15% to 30% per allele, and CYP4F23 carriers an increase of 5% to 10% (optional recommendations). For patients of African ancestry, genotype-guided dosing is used only if CYP2C95, 6, 8, and 11 have been tested; if they have not, dosing is clinical (moderate recommendation). When they are known, the algorithm dose is reduced by 15% to 30% per variant, and by 10% to 25% in African Americans carrying rs12777823 A. For any ancestry, an alternative anticoagulant is considered for CYP2C9 poor metabolizers (2/3, 3/*3) and for patients combining VKORC1 AA with reduced CYP2C9 function.

The International Warfarin Pharmacogenetics Consortium algorithm was derived in 4,043 patients and validated in 1,009. It predicted the stable dose more accurately than a clinical algorithm or a fixed-dose start. The gains were largest in the 46.2% of patients who needed 21 mg or less per week (dose correctly identified in 49.4% versus 33.3%) or 49 mg or more per week (24.8% versus 7.2%), the groups most at risk with standard dosing.

Genotype-guided dosing trials and guideline positions

The different comparator arms explain part of the difference in results: COAG compared two algorithms, while EU-PACT compared genotype-guided dosing with a fixed loading regimen. TTR, time in therapeutic range; VTE, venous thromboembolism.

The 2012 ACCP guideline recommended against routine pharmacogenetic testing to guide warfarin dosing (Grade 1B), a recommendation that preceded GIFT. The prescribing information states that genotype can inform the initial dose. In practice, genotype is used when it is already known (for example from preemptive panel testing) or when results can be available before the first doses. Testing after warfarin has been titrated adds little, because the INR response already reflects genotype. Genotype explains part of the dose but not drug interactions, diet, adherence, or illness, so INR monitoring is unchanged.

Managing a supratherapeutic INR without bleeding

A high INR in a patient without bleeding is managed by withholding warfarin and, at higher values, by giving low-dose oral vitamin K. The goal is a return to the therapeutic range, not full reversal. Over-correction exposes patients with mechanical valves or recent thrombosis to thrombosis, and large vitamin K doses produce resistance to warfarin for days afterward. Oral vitamin K acts within about 24 hours and is preferred over subcutaneous vitamin K, which is absorbed unpredictably (ACCP 2008, Grade 1A).

Thresholds differ slightly between documents (5.0 and 9.0 in ACCP 2008; 4.5 and 10 in ACCP 2012 and ASH 2018). The trial evidence is summarized below. Management of bleeding is covered on the anticoagulant bleeding, reversal, and periprocedural management page.

The evidence against routine vitamin K at moderate elevations comes from a randomized trial of 724 nonbleeding patients with an INR of 4.5 to 10.0. Oral vitamin K 1.25 mg lowered the INR faster than placebo (a mean fall of 2.8 versus 1.4 units the next day) but did not reduce bleeding over 90 days: any bleeding occurred in 15.8% with vitamin K and 16.3% with placebo, major bleeding in 2.5% and 1.1%, and thromboembolism in 1.1% and 0.8%. The trial did not include patients with bleeding, an INR above 10, or a need for urgent correction.

The cause of every high INR is identified before the dose is resumed: a new drug, antibiotic, or supplement, acute illness, reduced intake, a dosing error, or a change in tablet strength. When the cause is transient, the previous dose is resumed after the INR returns to range. When it persists, as with amiodarone, the weekly dose is reduced. After vitamin K, the INR is rechecked within 24 to 48 hours and again within a week, because warfarin requirements can be temporarily higher after vitamin K and then fall.

Bleeding on warfarin: risk, sites, and reversal

Major bleeding rates with warfarin in the atrial fibrillation trials were about 3% per year (3.09% per year in the ARISTOTLE warfarin arm), with intracranial hemorrhage the most serious event. The INR is the dominant modifiable determinant of intracranial bleeding. In a case-control study of outpatients on warfarin, the risk of intracerebral hemorrhage doubled with each 0.5 increase in the prothrombin time ratio. The risk of subdural hemorrhage rose sharply above a ratio corresponding to an INR of about 4.0. Older age, prior cerebrovascular disease, and a prosthetic valve were additional risk factors.

Bleeding risk scores such as HAS-BLED identify correctable factors (uncontrolled hypertension, labile INR, antiplatelet drugs or NSAIDs, alcohol excess). A high score is not by itself a reason to withhold anticoagulation from a patient at high thrombotic risk. The score is described with the stroke risk scores on the atrial fibrillation page.

Major or life-threatening bleeding is treated by withholding warfarin, achieving local hemostasis, transfusing as needed, and reversing with 4-factor prothrombin complex concentrate dosed by INR and body weight plus intravenous vitamin K. The vitamin K sustains the correction once the infused factors are cleared. Four-factor prothrombin complex concentrate is preferred over fresh frozen plasma because it corrects the INR faster with a small infusion volume (2023 AF guideline, Class 1, A; ASH 2018). Fresh frozen plasma is used only when prothrombin complex concentrate is unavailable, at volumes that risk circulatory overload. Recombinant factor VIIa, an option in the 2008 ACCP guideline, normalizes the INR through factor VII alone without replacing prothrombin or factor X, and 4-factor prothrombin complex concentrate has supplanted it. After major bleeding, oral anticoagulation is resumed within 90 days in patients with a continuing indication (ASH 2018, conditional), with the timing set by the bleeding site and thrombotic risk. Dosing of reversal agents, targets by site, and the timing of resumption after intracranial and gastrointestinal bleeding are covered on the anticoagulant bleeding and reversal page.

In acute ischemic stroke in a warfarin-treated patient, intravenous thrombolysis is contraindicated when the INR is above 1.7 (2026 American Heart Association/American Stroke Association acute ischemic stroke guideline), and endovascular thrombectomy remains available.

Interrupting warfarin for procedures and bridging

Many procedures with low bleeding risk are performed without stopping warfarin, with the INR in or near the lower part of the therapeutic range. Examples are dental extractions and cleaning (often with local hemostatic measures such as tranexamic acid mouthwash), cataract surgery, minor dermatologic procedures, and many endoscopic and electrophysiology procedures. In BRUISE CONTROL, 681 patients on warfarin with an annual thromboembolic risk of 5% or more undergoing pacemaker or defibrillator surgery were randomized to continued warfarin or heparin bridging. Clinically significant pocket hematoma occurred in 3.5% with continued warfarin and 16.0% with bridging. Continued warfarin is recommended over interruption with bridging for device surgery in these patients (2023 AF guideline, Class 1, A).

When interruption is needed, warfarin is stopped 5 days before the procedure, which allows the INR to fall below 1.5 in most patients, and an INR is measured the day before or the day of surgery. Warfarin is resumed within 24 hours after the procedure when hemostasis is secure (CHEST 2022). Because warfarin takes several days to re-establish its effect, early resumption carries little immediate bleeding risk.

In BRIDGE, among 1,884 patients with atrial fibrillation whose warfarin was interrupted for an elective procedure, no bridging was noninferior to dalteparin bridging for arterial thromboembolism (0.4% versus 0.3%) and caused less major bleeding (1.3% versus 3.2%). Bridging is therefore not used in atrial fibrillation without a mechanical valve or recent stroke (2023 AF guideline, Class 3: Harm for low-molecular-weight heparin), nor in low-to-moderate-risk venous thromboembolism (ASH 2018, strong recommendation against). It is reserved for patients at very high thrombotic risk: a mechanical mitral valve, an older-generation mechanical valve, a mechanical aortic valve with risk factors, a stroke or systemic embolism within about 3 months, or venous thromboembolism within 3 months. Bridging for valves is covered on the prosthetic heart valves page, and the full periprocedural approach on the periprocedural management page.

Switching between warfarin and DOACs

Switching requires attention to the INR because warfarin acts slowly and DOACs act within hours, and because DOACs themselves raise the INR.

From the current US prescribing information for each drug. The ASH 2018 guideline suggests overlapping the DOAC with the vitamin K antagonist until the INR is therapeutic, over bridging with low-molecular-weight or unfractionated heparin (conditional). When a DOAC and warfarin overlap, the INR is drawn just before the next DOAC dose, when DOAC interference is lowest.

The transition from a DOAC to warfarin carries the higher risk of a gap in protection, because warfarin takes days to act and the INR is unreliable while the DOAC is present. The patient with a new mechanical valve, newly diagnosed moderate or severe rheumatic mitral stenosis, or newly confirmed triple-positive antiphospholipid syndrome while taking a DOAC is moved to warfarin with continuous coverage until the INR is therapeutic.

Nonbleeding adverse effects: skin necrosis, venous limb gangrene, calciphylaxis, purple toes, and nephropathy

From the US prescribing information (sections 5 and 6) and the cited studies. HIT, heparin-induced thrombocytopenia.

Skin necrosis. Lesions begin as painful erythema or purpura in areas of abundant subcutaneous fat (breast, thigh, buttock, abdomen) and progress to hemorrhagic bullae and full-thickness necrosis. Histology shows thrombi in dermal and subcutaneous venules without inflammation. The syndrome can occur in patients without a known deficiency. Prevention depends on avoiding large loading doses in patients with known deficiency and on parenteral overlap in patients with acute thrombosis.

Venous limb gangrene. Among 158 patients with heparin-induced thrombocytopenia, all 8 who developed venous limb gangrene were receiving warfarin, compared with 3 of 10 with limb arterial thrombosis. Their median INR was 5.8, compared with 3.1 in patients with deep vein thrombosis without gangrene, reflecting factor VII depletion. Their plasma showed a higher ratio of thrombin-antithrombin complexes to protein C activity. The ASH 2018 HIT guideline and the prescribing information advise against starting warfarin until the platelet count has recovered, usually to 150 × 10⁹/L or more, and overlap with a nonheparin anticoagulant continues for at least 5 days and until the INR is therapeutic. If warfarin was already started when HIT is recognized, vitamin K is given. Agent selection and the 4Ts score are covered on the parenteral anticoagulants and heparin-induced thrombocytopenia page.

Calciphylaxis. Matrix Gla protein inhibits vascular calcification only after vitamin K-dependent carboxylation. In 20 patients receiving hemodialysis with calciphylaxis and 20 controls matched for age, sex, race, and warfarin use, the carboxylated fraction of matrix Gla protein was lower in patients with calciphylaxis (0.58 versus 0.69). Each 0.1 reduction was associated with a more than twofold increase in risk, and vitamin K deficiency was associated with lower carboxylation. Warfarin blocks this carboxylation and is a recognized risk factor; the prescribing information lists calciphylaxis as a warning in patients with and without kidney failure.

Nephropathy. Anticoagulant-related nephropathy is acute kidney injury caused by glomerular hemorrhage with obstruction of tubules by red cell casts during excessive anticoagulation. In a cohort of 15,258 patients starting warfarin, 4,006 had an INR above 3.0 with a creatinine measured at the same time. A presumptive diagnosis (creatinine rise of more than 0.3 mg/dL within a week of an INR above 3.0 without recorded hemorrhage) was made in 20.5% of these patients, 33.0% of those with chronic kidney disease, and 16.5% of those without it. One-year mortality was 31.1% with and 18.9% without the diagnosis. The prescribing information advises more frequent INR monitoring in patients with kidney disease.

Special populations: kidney and liver disease, older adults, cancer, obesity, and bariatric surgery

Chronic kidney disease and dialysis. Warfarin is not renally cleared, and the prescribing information requires no dose adjustment, but kidney disease is associated with lower dose requirements, more labile INR values, more bleeding, and the risks of nephropathy and calciphylaxis. In atrial fibrillation, the 2023 guideline recommends warfarin or, preferably, a DOAC at the evidence-based dose in stage 3 chronic kidney disease (Class 1). For stage 4, it considers warfarin or a labeled-dose DOAC reasonable (Class 2a). For kidney failure or dialysis, warfarin with an INR of 2.0 to 3.0 or an evidence-based dose of apixaban might be reasonable (Class 2b). No adequately powered randomized trial has shown net benefit of any anticoagulant over none in patients receiving dialysis.

Liver disease. Cirrhosis lowers both procoagulant and anticoagulant factors, and the baseline INR does not measure bleeding risk. Warfarin is sensitive and labile in this setting, and the target INR is interpreted against an elevated baseline. Warfarin and low-molecular-weight heparin have the most data for portal vein thrombosis in cirrhosis, and DOACs are avoided in Child-Pugh class C. Chromogenic factor X testing is used in some centers when the baseline INR is prolonged.

Older adults. Older patients need lower doses, have more interacting drugs and comorbidities, and have a higher bleeding risk at any INR. A starting dose of 5 mg or less is recommended (ACCP 2008, Grade 1C). Age alone does not change the INR target, because lower targets increase stroke severity and mortality in atrial fibrillation. The absolute benefit of anticoagulation increases with age. Cognitive impairment without a caregiver to supervise dosing is a practical contraindication.

Cancer. Cancer increases both thrombosis and bleeding. Warfarin is less effective than low-molecular-weight heparin for cancer-associated venous thromboembolism, and interactions with chemotherapy (fluorouracil, capecitabine, and many targeted agents) make the INR labile. Low-molecular-weight heparin and factor Xa inhibitors are preferred, as covered on the pulmonary embolism page.

Obesity and body size. Warfarin is titrated to the INR, so obesity does not change the monitoring strategy. Higher body weight and height are associated with higher maintenance doses, and weight is included in pharmacogenetic and clinical algorithms. Warfarin is often used when DOAC dosing is uncertain at extremes of body weight, although apixaban and rivaroxaban are used in severe obesity with supporting data.

Bariatric surgery and malabsorption. After bariatric surgery, warfarin requirements often fall in the first weeks with reduced intake and then rise. Because the INR measures the warfarin effect, warfarin is used when absorption of a DOAC is uncertain after gastric bypass or extensive small bowel resection.

Mechanical circulatory support. Patients with durable left ventricular assist devices receive warfarin according to device-specific protocols, with INR targets set by the implanting program. Pump thrombosis is associated with subtherapeutic INR values, and bleeding (gastrointestinal angiodysplasia from acquired von Willebrand syndrome) with high values.

Pregnancy and lactation

Warfarin crosses the placenta. Exposure in the first trimester, chiefly weeks 6 to 12, causes warfarin embryopathy (nasal hypoplasia and stippled epiphyses) in about 5% of exposed pregnancies in the data cited by the prescribing information, and exposure at any stage can cause fetal intracranial hemorrhage and fetal loss. Warfarin is contraindicated in pregnancy except in women with mechanical heart valves at high thromboembolic risk. For venous thromboembolism and most other indications, low-molecular-weight heparin is used throughout pregnancy and warfarin is avoided.

In women with mechanical valves, fetal risk depends on dose. The 2020 ACC/AHA guideline considers continuation of warfarin through all trimesters reasonable when the dose required is 5 mg daily or less (Class 2a), with dose-adjusted low-molecular-weight heparin in the first trimester as an alternative. When more than 5 mg daily is required, dose-adjusted low-molecular-weight heparin with anti-Xa monitoring in the first trimester, followed by warfarin, is reasonable (Class 2a). Warfarin is changed to heparin at least 1 week before planned delivery. Detailed recommendations are on the prosthetic heart valves page.

Warfarin was not detected in human milk in a limited study, and breastfeeding is compatible with warfarin, with observation of the infant for bruising or bleeding. Warfarin is the preferred oral anticoagulant after delivery in women who need long-term anticoagulation and are breastfeeding, because DOAC safety in lactation is not established.

Missed doses, patient education, and adherence

Because the anticoagulant effect persists beyond 24 hours, a missed dose is taken as soon as possible on the same day. Doses are not doubled the next day to make up for a missed dose. Patients record missed or extra doses so that the next INR can be interpreted. A missed dose in a patient with a mechanical mitral valve is a reason for an earlier INR, not for an extra dose.

Education at initiation and at each change covers:

• the indication, target INR, and planned duration
• the meaning of the INR and the testing schedule
• tablet strengths and colors and the weekly schedule
• consistent vitamin K intake and alcohol use
• the need to report every new prescription, antibiotic, over-the-counter analgesic, and supplement
• signs of bleeding (melena, hematuria, unusual bruising, prolonged nosebleeds, severe headache) and of thrombosis
• avoidance of contact sports and activities with serious injury risk
• pregnancy planning and contraception for women of reproductive potential
• notification of all clinicians and dentists before procedures

A wallet card or medical alert identification lists warfarin and the indication.

The ASH 2018 guideline advises against adherence aids studied in randomized trials, namely daily lottery incentives, electronic reminders, and visual medication schedules (very low certainty). It suggests care by a specialized anticoagulation management service and supplementary patient education in addition to basic education (both conditional, very low certainty). The 2012 ACCP guideline advises coordinated, systematic management of warfarin, with patient education, INR testing, tracking, follow-up, and clear communication of results and dosing decisions (best-practice statement).

Sources of error in warfarin management

• Treating an early INR as protection. An INR of 2.0 on day 2 or 3 reflects factor VII depletion. Stopping heparin before 5 days of overlap in acute thrombosis leaves the patient unprotected while protein C is low.
• Applying a DOAC regimen to a warfarin indication. A mechanical valve, rheumatic or moderate or greater mitral stenosis with atrial fibrillation, and triple-positive antiphospholipid syndrome require warfarin. Each is excluded by history, echocardiography, or antibody testing before a DOAC is chosen.
• Giving large loading doses to sensitive patients. Older, malnourished, or postoperative patients and those taking amiodarone or with a sensitive genotype develop very high INR values on 10-mg starts.
• Changing the dose on every INR. Adjustments made before a previous change has reached steady state produce oscillation. A single slightly out-of-range value in a stable patient does not require a dose change.
• Missing the delayed amiodarone interaction. The INR rises over several weeks after amiodarone starts and remains affected for months after it stops.
• Overlooking antibiotics, acetaminophen, and supplements. Regular acetaminophen above about 2 g daily, any antibiotic, and botanicals change the INR and are often not reported.
• Reversing a nonbleeding high INR in full. Vitamin K for an INR of 4.5 to 10 without bleeding does not reduce bleeding, and large doses cause warfarin resistance and thrombotic exposure in valve patients.
• Bridging every interruption. Bridging in atrial fibrillation without a mechanical valve or recent stroke increases bleeding without reducing thromboembolism.
• Reading the INR during DOAC or argatroban overlap. These drugs raise the INR, so timing of the sample and agent-specific transition rules apply.
• Trusting a point-of-care INR in lupus anticoagulant. Phospholipid-dependent interference can overestimate the INR; laboratory confirmation or a chromogenic factor X assay is used.
• Starting warfarin in acute heparin-induced thrombocytopenia. Warfarin before platelet recovery causes venous limb gangrene.
• Using genotype after titration. Once the INR response is known, genotype adds little to dose decisions.

References

1. Jantoven (warfarin sodium tablets) prescribing information. Upsher-Smith Laboratories; label revised 2026. DailyMed set ID 19a69a72-ac5d-45d5-a94d-a5aaecbe4730. dailymed.nlm.nih.gov
2. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians evidence-based clinical practice guidelines (8th edition). Chest. 2008;133(6 Suppl):160S–198S. doi:10.1378/chest.08-0670
3. Ageno W, Gallus AS, Wittkowsky A, Crowther M, Hylek EM, Palareti G. Oral anticoagulant therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2 Suppl):e44S–e88S. doi:10.1378/chest.11-2292
4. Holbrook A, Schulman S, Witt DM, et al. Evidence-based management of anticoagulant therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2 Suppl):e152S–e184S. doi:10.1378/chest.11-2295
5. Witt DM, Nieuwlaat R, Clark NP, et al. American Society of Hematology 2018 guidelines for management of venous thromboembolism: optimal management of anticoagulation therapy. Blood Adv. 2018;2:3257–3291. doi:10.1182/bloodadvances.2018024893
6. Otto CM, Nishimura RA, Bonow RO, et al. 2020 ACC/AHA guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2021;143:e72–e227. Corrections: Circulation. 2021;143:e229; 2023;148:e8; 2023;148:e185; 2024;150:e267. doi:10.1161/CIR.0000000000000923
7. Joglar JA, Chung MK, Armbruster AL, et al. 2023 ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2024;149:e1–e156. Corrections: Circulation. 2024;149:e167; 2024;149:e936; 2024;149:e1413. doi:10.1161/CIR.0000000000001193
8. Van Gelder IC, Rienstra M, Bunting KV, et al. 2024 ESC guidelines for the management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2024;45:3314–3414. doi:10.1093/eurheartj/ehae176
9. Praz F, Borger MA, Lanz J, et al. 2025 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J. 2025;46:4635–4736. doi:10.1093/eurheartj/ehaf194
10. Stevens SM, Woller SC, Baumann Kreuziger L, et al. Antithrombotic therapy for VTE disease: second update of the CHEST guideline and expert panel report. Chest. 2021;160:e545–e608. doi:10.1016/j.chest.2021.07.055
11. Douketis JD, Spyropoulos AC, Murad MH, et al. Perioperative management of antithrombotic therapy: an American College of Chest Physicians clinical practice guideline. Chest. 2022;162:e207–e243. doi:10.1016/j.chest.2022.07.025
12. Tomaselli GF, Mahaffey KW, Cuker A, et al. 2020 ACC expert consensus decision pathway on management of bleeding in patients on oral anticoagulants: a report of the American College of Cardiology Solution Set Oversight Committee. J Am Coll Cardiol. 2020;76:594–622. doi:10.1016/j.jacc.2020.04.053
13. Johnson JA, Caudle KE, Gong L, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for pharmacogenetics-guided warfarin dosing: 2017 update. Clin Pharmacol Ther. 2017;102:397–404. doi:10.1002/cpt.668
14. International Warfarin Pharmacogenetics Consortium; Klein TE, Altman RB, Eriksson N, et al. Estimation of the warfarin dose with clinical and pharmacogenetic data. N Engl J Med. 2009;360:753–764. doi:10.1056/NEJMoa0809329
15. Kimmel SE, French B, Kasner SE, et al. A pharmacogenetic versus a clinical algorithm for warfarin dosing. N Engl J Med. 2013;369:2283–2293. doi:10.1056/NEJMoa1310669
16. Pirmohamed M, Burnside G, Eriksson N, et al. A randomized trial of genotype-guided dosing of warfarin. N Engl J Med. 2013;369:2294–2303. doi:10.1056/NEJMoa1311386
17. Gage BF, Bass AR, Lin H, et al. Effect of genotype-guided warfarin dosing on clinical events and anticoagulation control among patients undergoing hip or knee arthroplasty: the GIFT randomized clinical trial. JAMA. 2017;318:1115–1124. doi:10.1001/jama.2017.11469
18. Kovacs MJ, Rodger M, Anderson DR, et al. Comparison of 10-mg and 5-mg warfarin initiation nomograms together with low-molecular-weight heparin for outpatient treatment of acute venous thromboembolism: a randomized, double-blind, controlled trial. Ann Intern Med. 2003;138:714–719. doi:10.7326/0003-4819-138-9-200305060-00007
19. Crowther MA, Ginsberg JB, Kearon C, et al. A randomized trial comparing 5-mg and 10-mg warfarin loading doses. Arch Intern Med. 1999;159:46–48. doi:10.1001/archinte.159.1.46
20. Pengo V, Biasiolo A, Pegoraro C. A simple scheme to initiate oral anticoagulant treatment in outpatients with nonrheumatic atrial fibrillation. Am J Cardiol. 2001;88:1214–1216. doi:10.1016/S0002-9149(01)02069-0
21. Van Spall HG, Wallentin L, Yusuf S, et al. Variation in warfarin dose adjustment practice is responsible for differences in the quality of anticoagulation control between centers and countries: an analysis of patients receiving warfarin in the Randomized Evaluation of Long-term Anticoagulation Therapy (RE-LY) trial. Circulation. 2012;126:2309–2316. doi:10.1161/CIRCULATIONAHA.112.101808
22. Rosendaal FR, Cannegieter SC, van der Meer FJ, Briët E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost. 1993;69:236–239. doi:10.1055/s-0038-1651587
23. Apostolakis S, Sullivan RM, Olshansky B, Lip GYH. Factors affecting quality of anticoagulation control among patients with atrial fibrillation on warfarin: the SAMe-TT₂R₂ score. Chest. 2013;144:1555–1563. doi:10.1378/chest.13-0054
24. Heneghan C, Ward A, Perera R, et al. Self-monitoring of oral anticoagulation: systematic review and meta-analysis of individual patient data. Lancet. 2012;379:322–334. doi:10.1016/S0140-6736(11)61294-4
25. Matchar DB, Jacobson A, Dolor R, et al. Effect of home testing of international normalized ratio on clinical events. N Engl J Med. 2010;363:1608–1620. doi:10.1056/NEJMoa1002617
26. Holbrook AM, Pereira JA, Labiris R, et al. Systematic overview of warfarin and its drug and food interactions. Arch Intern Med. 2005;165:1095–1106. doi:10.1001/archinte.165.10.1095
27. Sanoski CA, Bauman JL. Clinical observations with the amiodarone/warfarin interaction: dosing relationships with long-term therapy. Chest. 2002;121:19–23. doi:10.1378/chest.121.1.19
28. Hylek EM, Heiman H, Skates SJ, Sheehan MA, Singer DE. Acetaminophen and other risk factors for excessive warfarin anticoagulation. JAMA. 1998;279:657–662. doi:10.1001/jama.279.9.657
29. Hylek EM, Go AS, Chang Y, et al. Effect of intensity of oral anticoagulation on stroke severity and mortality in atrial fibrillation. N Engl J Med. 2003;349:1019–1026. doi:10.1056/NEJMoa022913
30. Hylek EM, Singer DE. Risk factors for intracranial hemorrhage in outpatients taking warfarin. Ann Intern Med. 1994;120:897–902. doi:10.7326/0003-4819-120-11-199406010-00001
31. Crowther MA, Ageno W, Garcia D, et al. Oral vitamin K versus placebo to correct excessive anticoagulation in patients receiving warfarin: a randomized trial. Ann Intern Med. 2009;150:293–300. doi:10.7326/0003-4819-150-5-200903030-00005
32. Connolly SJ, Karthikeyan G, Ntsekhe M, et al. Rivaroxaban in rheumatic heart disease-associated atrial fibrillation. N Engl J Med. 2022;387:978–988. doi:10.1056/NEJMoa2209051
33. Eikelboom JW, Connolly SJ, Brueckmann M, et al. Dabigatran versus warfarin in patients with mechanical heart valves. N Engl J Med. 2013;369:1206–1214. doi:10.1056/NEJMoa1300615
34. Wang TY, Svensson LG, Wen J, et al. Apixaban or warfarin in patients with an On-X mechanical aortic valve. NEJM Evid. 2023;2(7):EVIDoa2300067. doi:10.1056/EVIDoa2300067
35. Pengo V, Denas G, Zoppellaro G, et al. Rivaroxaban vs warfarin in high-risk patients with antiphospholipid syndrome. Blood. 2018;132:1365–1371. doi:10.1182/blood-2018-04-848333
36. Crowther MA, Ginsberg JS, Julian J, et al. A comparison of two intensities of warfarin for the prevention of recurrent thrombosis in patients with the antiphospholipid antibody syndrome. N Engl J Med. 2003;349:1133–1138. doi:10.1056/NEJMoa035241
37. Tektonidou MG, Andreoli L, Limper M, et al. EULAR recommendations for the management of antiphospholipid syndrome in adults. Ann Rheum Dis. 2019;78:1296–1304. doi:10.1136/annrheumdis-2019-215213
38. Moll S, Ortel TL. Monitoring warfarin therapy in patients with lupus anticoagulants. Ann Intern Med. 1997;127:177–185. doi:10.7326/0003-4819-127-3-199708010-00001
39. Levine GN, McEvoy JW, Fang JC, et al. Management of patients at risk for and with left ventricular thrombus: a scientific statement from the American Heart Association. Circulation. 2022;146:e205–e223. doi:10.1161/CIR.0000000000001092
40. Douketis JD, Spyropoulos AC, Kaatz S, et al. Perioperative bridging anticoagulation in patients with atrial fibrillation. N Engl J Med. 2015;373:823–833. doi:10.1056/NEJMoa1501035
41. Birnie DH, Healey JS, Wells GA, et al. Pacemaker or defibrillator surgery without interruption of anticoagulation. N Engl J Med. 2013;368:2084–2093. doi:10.1056/NEJMoa1302946
42. Cuker A, Arepally GM, Chong BH, et al. American Society of Hematology 2018 guidelines for management of venous thromboembolism: heparin-induced thrombocytopenia. Blood Adv. 2018;2:3360–3392. doi:10.1182/bloodadvances.2018024489
43. Warkentin TE, Elavathil LJ, Hayward CP, Johnston MA, Russett JI, Kelton JG. The pathogenesis of venous limb gangrene associated with heparin-induced thrombocytopenia. Ann Intern Med. 1997;127:804–812. doi:10.7326/0003-4819-127-9-199711010-00005
44. Brodsky SV, Nadasdy T, Rovin BH, et al. Warfarin-related nephropathy occurs in patients with and without chronic kidney disease and is associated with an increased mortality rate. Kidney Int. 2011;80:181–189. doi:10.1038/ki.2011.44
45. Nigwekar SU, Bloch DB, Nazarian RM, et al. Vitamin K-dependent carboxylation of matrix Gla protein influences the risk of calciphylaxis. J Am Soc Nephrol. 2017;28:1717–1722. doi:10.1681/ASN.2016060651
46. Wittkowsky AK. Dietary supplements, herbs and oral anticoagulants: the nature of the evidence. J Thromb Thrombolysis. 2008;25:72–77. doi:10.1007/s11239-007-0110-0
47. Harrison L, Johnston M, Massicotte MP, Crowther M, Moffat K, Hirsh J. Comparison of 5-mg and 10-mg loading doses in initiation of warfarin therapy. Ann Intern Med. 1997;126:133–136. doi:10.7326/0003-4819-126-2-199701150-00006
48. Prabhakaran S, Gonzalez NR, Zachrison KS, et al. 2026 guideline for the early management of patients with acute ischemic stroke: a guideline from the American Heart Association/American Stroke Association. Stroke. 2026;57:e316–e436. Correction: Stroke. 2026;57:e461–e467. doi:10.1161/STR.0000000000000513
49. Puskas J, Gerdisch M, Nichols D, et al. Reduced anticoagulation after mechanical aortic valve replacement: interim results from the Prospective Randomized On-X Valve Anticoagulation Clinical Trial randomized Food and Drug Administration investigational device exemption trial. J Thorac Cardiovasc Surg. 2014;147:1202–1210. doi:10.1016/j.jtcvs.2014.01.004
50. Puskas JD, Gerdisch M, Nichols D, et al. Anticoagulation and antiplatelet strategies after On-X mechanical aortic valve replacement. J Am Coll Cardiol. 2018;71:2717–2726. doi:10.1016/j.jacc.2018.03.535