Warfarin is the most widely used oral anticoagulant in Australia. Although it can prevent thrombosis, it can cause life-threatening haemorrhages. Patients taking warfarin should have their INR measured regularly. More frequent tests are needed when patients start, stop or alter the dose of their other medications. Educating the patients about warfarin helps them to take their treatment correctly. They should report any abnormal bleeding and have their INR measured. A very high INR is an indication for admission to hospital to have the effects of warfarin controlled.
Warfarin is one of the commonest causes of death related to prescription drugs, but when used appropriately it is one of the most beneficial drugs. As the population ages and more trials show its benefits, more and older patients will be started on warfarin. Its pharmacology is complicated and many factors need to be considered in the optimal management of each patient.
Risks and benefits of warfarin
There are many indications for warfarin therapy (Table 1). The decision to start warfarin depends on an assessment of each patient's balance between the harmful effects and the benefits of anticoagulation.
The common indications for warfarin therapy
Supported with good evidence
Supported but limited evidence
Little or no supporting evidence
The major adverse effect of warfarin is an increased bleeding tendency and many factors can increase the risk (Table 2).1A patient's risk of bleeding is greatest in the first few months after starting warfarin. Bleeding complications occur in 3-10% of patients on warfarin per year, but most bleeds are minor. Although the bleeding risk increases as the INR (International Normalised Ratio) increases, 50% of bleeding episodes occur while the INR is less than 4.0.
Risk factors for major bleeding in patients on warfarin
Marked increase in risk
Moderate increase in risk
Age is one of the strongest risk factors for bleeding. In one study, the annual risk of major bleeding was 2.9% for patients older than 70 years, while no major bleeds occurred in patients under 50 years old.2Bleeding in this report was called `major' if it was fatal, was intracranial, retroperitoneal or involved a joint, required surgery, led to a haemoglobin fall of 2 g/dL or more, and/or required the transfusion of two or more units of blood.
Warfarin in pregnancy is teratogenic and causes peripartum bleeding in mother and child, so it is generally contraindicated in pregnancy. There may be a place for mid-trimester warfarin in pregnant women with prosthetic heart valves, but this choice should be made only after a full discussion of the implications with the patient. Unfractionated heparin and low molecular weight heparin are alternatives that do not cross the placenta.
Patients who have a poor understanding of the indications and potential adverse effects are more likely to be non-compliant than those who receive education about warfarin.3Patients should be encouraged to:
– report any signs of bleeding while on warfarin
– have more frequent measurements of their INR when starting or stopping other medications, either prescribed or complementary
– keep a written record of INR results and warfarin dosage
– remain with one or other of the currently available brands of warfarin (Coumadin or Marevan), as these have not been formally shown to be bioequivalent and are therefore not interchangeable.
Booklets aimed at patient education are available from all pharmacies. These are a useful supplement to the doctor's advice.
Women of childbearing age should be informed of the potential dangers of warfarin in pregnancy. Pre-conception counselling is needed for women on long-term treatment.
Trials of patient-based self-monitoring and dose-adjustment of warfarin therapy are under way, but currently this cannot be recommended.
Starting warfarin therapy
Patients with acute thromboembolic events, such as deep vein thrombosis, pulmonary embolism or embolic stroke, should be given heparin or low molecular weight heparin when starting warfarin. The heparin or low molecular weight heparin can be ceased after a minimum period of five days of combined therapy with warfarin and after the INR has been in the therapeutic range for 48 hours. Patients at less immediate risk, such as patients in stable atrial fibrillation without embolic events, may be safely started on warfarin without concurrent heparin.
Several trials have shown that low dose induction regimens cause fewer episodes of over-anticoagulation, with only minimal delay in the time to achieve the target INR. One protocol, validated specifically in elderly patients, is given in Table 3.4Loading regimens of5 mg warfarin per day in all patients have also been reported.5Previously, a modified Fennerty's protocol, using loading doses of 10 mg, 10 mg and 5 mg on the first three days, has been the most commonly used nomogram for starting warfarin. However, this protocol can lead to significant over-anticoagulation, particularly in the elderly.4With all induction protocols, a baseline INR is measured before starting warfarin, and the INR is measured each day until doses are stabilised.
A dosage schedule for starting warfarin therapy, validated in elderly patients
|Dose adjustment after day 4 depends on clinical judgement based on the pattern of INR.|
The table is slightly modified from the table in reference 4.
Maintaining warfarin therapy
The target INR varies for different clinical situations.6For patients with mechanical prosthetic heart valves, the target INR is usually 2.5-3.5. For almost all other conditions, including deep vein thrombosis and pulmonary embolism, atrial fibrillation and tissue valve replacements, the target INR is usually 2.0-3.0. Patients with recurrent thromboembolic events while anticoagulated, and patients with cancer or antiphospholipid syndrome may benefit from a higher INR, but this approach should be undertaken with specialist advice.
The frequency of monitoring INR once the dose is stabilised should be determined by the clinical situation. Initially, patients will require a few tests each week, but this can be gradually decreased to once a week or once a fortnight if the INR is stable. Patients who have a very stable INR, no interacting medications and low bleeding risk, may only need to have a test once every four to six weeks.
Many hospitals have anticoagulation clinics for managing patients on warfarin, although most patients can be safely managed in the community. However, clinical trials suggest that patients managed at an anticoagulant clinic spend more time with a therapeutic INR and have a lower rate of major haemorrhage, compared to `best usual care'.7Patients most likely to benefit from anticoagulation clinics include those with comorbidities, those with unstable INRs and those very sensitive to warfarin (requiring daily doses of 1 mg or less).
Interactions with warfarin
Warfarin is particularly prone to interactions with other drugs, herbal medicines8and dietary factors (Table 4). Some interactions involve the cytochrome P450 system in the liver. Many interactions are unpredictable, so the INR should be tested more frequently after starting a new medication and similarly when stopping a medication or changing the dose. It takes about five days for enzyme induction to take place, so that an INR measured about one week after a change in medication should reflect any interaction. In one study, recent antibiotic use was the second greatest risk factor (after age) for over-anticoagulation.9
Some of the important interactions of warfarin
|Increased effect of warfarin (INR higher)||Decreased effect of warfarin (INR lower)||Potentiate bleeding risk because of antiplatelet effect||Potentiate bleeding risk by effects on gastric mucosa|
– Antibiotics (sulfonamides,cerythromycin and other macrolides, metronidazole)
– Antifungals (itraconazole, fluconazole, ketoconazole)
– Selective serotonin reuptake inhibitors (especially fluvoxamine, fluoxetine)
– Quinine and quinidine
– COX-2 inhibitors (celecoxib, rofecoxib)
– Antiepileptics (carbamazepine, phenytoin, barbiturates)
– Rifampicin, rifabutin
– Non-steroidal anti-inflammatory drugs(except COX-2 inhibitors)
– Non-steroidal anti-inflammatory drugs
– Dong quai
– St John's wort
Alcohol in small to moderate amounts probably has little effect on warfarin metabolism. In heavy drinkers, however, factors such as increased falls, alcohol-induced gastritis, poor diet and poor compliance potentiate the risk of bleeding.
The amount of vitamin K in the diet partly determines the sensitivity to warfarin. A diet high in vitamin K reduces the response. This is important to consider in situations when diet changes, such as during illness, travel, fad diets, hospitalisation and postoperatively. Foods high in vitamin K include green tea, turnips, avocados, brussel sprouts, broccoli and green leafy vegetables (e.g. lettuce, cabbage). It takes a very large daily intake of `greens' to influence the INR. In any case, a consistently sustained diet will minimise this potential source of fluctuating results.
Management of over-anticoagulation
Over-anticoagulation increases the risk of haemorrhage. The first step in managing this problem is to identify the cause. Common causes include starting or stopping an interacting medication, deteriorating liver function, and patient error (such as taking the wrong dose or confusing different strength tablets). Many of these causes are preventable.
The approach to a raised INR should be individualised, paying attention to the indication for the warfarin, the patient's risk of bleeding and whether it is safe to continue therapy at all. Some patients need to be admitted to hospital, while others just need to miss a dose of warfarin.
Guidelines for managing over-anticoagulation (Table 5) are based on the recently published recommendations from the Australasian Society of Thrombosis and Haemostasis.10The half-life of vitamin K is shorter than that of warfarin, so the INR may rebound 24-48 hours after giving vitamin K. The intravenous preparation of vitamin K can be administered orally or subcutaneously with equal efficacy, and these routes are usually safer and more convenient in patients who are not actively bleeding.
A protocol for managing over-anticoagulation9
Withhold warfarin and measure INR next day
Vitamin K 1-2.5 mg subcutaneously or orally*
Recheck INR next day
Vitamin K 5 mg IV or subcutaneously*
Fresh frozen plasma 2 Units. This may be given with a factor II, VII, IX concentrate
Recheck INR after 6-8 hours and then daily for 3 days
|* The intravenous preparation may be given orally or subcutaneously with safety and efficacy. Not all community pharmacies have the intravenous formulation of vitamin K and it may be worth keeping a supply in the practice rooms. (Avoid intramuscular injections of vitamin K to prevent local injection site bleeding which also reduces bioavailability.)|
Fresh frozen plasma and concentrates of clotting factors are blood products and may carry a small risk of viral contamination.
The following statements are either true or false.
Click anywhere on the panel for the answers.
1. The risk of warfarin causing bleeding is the same in all age groups.
2. The INR of a patient taking warfarin may be altered by a change in diet.
- Beyth RJ, Romano PS. Major bleeding after hospitalization for deep-venous thrombosis. Am J Med 1999;107:414-24. .
- Palareti G, Leali N, Coccheri S, Poggi M, D'Angelo A, Manotti C. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Lancet 1996;348:423-8. .
- Arnsten JH, Gelfand JM. Determinants of compliance with anticoagulation: A case-control study. Am J Med 1997;103:11-7. .
- Gedge J, Orme S, Hampton KK, Hendra TJ. A comparison of a low-dose warfarin induction regimen with the modified Fennerty regimen in elderly inpatients. Age Ageing 2000;29:31-4. .
- Harrison L, Massicotte MP, Moffat K. Comparison of 5-mg and 10-mg loading doses in initiation of warfarin therapy. Ann Intern Med 1997;126:133-6. .
- Fifth ACCP Consensus Conference on Antithrombotic Therapy. Chest 1998;114 Suppl 5:439S-769S. .
- Comparison of an anticoagulation clinic with usual medical care: anticoagulation control, patient outcomes, and health care costs. Arch Intern Med 1998;158:1641-7. .
- Herb-drug interactions. Lancet 2000;355:134-8. .
- Panneerselvam S, Lefort W. Analysis of risk factors for over-anticoagulation in patients receiving long-term warfarin. Br J Haematol 1998;103:422-4. .
- Consensus guidelines for warfarin therapy. Recommendations from the Australasian Society of Thrombosis and Haemostasis. Med J Aust 2000;172:600-5 .