View / Download
pdf version of this article
Mechanisms include inhibition or induction of warfarin metabolism, additive or synergistic antiplatelet effects (e.g.
aspirin) and changing vitamin K status (e.g. dietary changes and antibiotics). Individual susceptibility may increase
the risk of an interaction in some people. For many drugs the supporting evidence for an interaction is relatively weak
and based on poor quality case reports. However, the consequences of an interaction are potentially severe so any change
(starting or stopping) in drug therapy, diet or herbal/dietary supplement intake should be checked for the potential to
alter anticoagulant control.
Rather than an extensive list of individual drugs we have focused on the major drug classes (plus some individual drugs
which don’t fit) to give some indication of the relative potential for interactions. Hopefully this will be easy to read
and digest and prompt the need for further advice on management when necessary. As this ‘list’ is not totally comprehensive
or inclusive our advice is to check any agents’s potential to interact with warfarin, especially if you are unfamiliar
with the combination.
Acid regulating drugs
Cimetidine inhibits the metabolism of warfarin and increases its anticoagulant effect. Ranitidine appears not to interact
but there have been isolated cases of bleeding. There have been some reports of bleeding with proton pump inhibitors (PPIs)
and warfarin but an interaction is relatively uncommon. An awareness of the slight possibility of bleeding along with
the usual monitoring of the INR are required when PPIs are given with warfarin.
Anticoagulant control can be affected by changes in alcohol consumption. Poor diet and liver damage in alcoholics may
also be factors.
Most patients are not affected. Clinically important increases in INR are rarely reported.
Monitor the INR when allopurinol is first added.
Most analgesics can increase the bleeding risk when given with warfarin but by different mechanisms. Aspirin and NSAIDs
have antiplatelet effects so an increased risk of bleeding can occur without an increase in the INR. Warfarin can increase
the severity of a GI bleed caused by aspirin, NSAIDs and COX-2 inhibitors. Other than occasional case reports there is
little evidence that NSAIDs increase the INR in people taking warfarin. There is some evidence that the COX-2 inhibitors
(e.g. celecoxib, etoricoxib) can cause a significant increase in the INR in some people on warfarin, especially the elderly.
Consider monitoring the INR if a COX-2 inhibitor is added to warfarin.
Recent case reports indicate that tramadol can inhibit the metabolism of warfarin and increase its anticoagulant effect.
Regular use of paracetamol can increase the effect of warfarin in some patients. Occasional doses of paracetamol for
two to three days are unlikely to have a significant effect, but regular use for a week or more may increase the INR in
some patients. Monitor the INR in people starting or stopping regular use of paracetamol.
Amiodarone inhibits the metabolism of warfarin and this interaction occurs in most patients. The onset may be slow (about
two weeks) and it may persist for weeks or even months after the amiodarone has been stopped. The anticoagulant effect
of warfarin is possibly enhanced by quinidine.
Most antibiotics have been reported to alter the anticoagulant effect of warfarin and the usual mechanism proposed is
reduction of vitamin K synthesis by gut flora. Monitoring is recommended when an antibiotic is started or stopped. Clinically
significant interactions appear to be more likely with ciprofloxacin, norfloxacin, macrolides (e.g. erythromycin) and
metronidazole, where enzyme inhibition is also possibly involved.
Rifampicin is a potent enzyme inducer and can reduce the effect of warfarin.
Selective serotonin reuptake inhibitors (SSRIs) have antiplatelet effects and as well as increasing the risk of bleeding
per se they can enhance the risk of bleeding with warfarin without increasing the INR. There are also isolated case reports
of SSRIs causing an increase in INR. Advise people to report signs of bleeding and consider monitoring the INR when an
SSRI is added.
The effects of warfarin are occasionally increased by venlafaxine, sometimes with an increase in the INR.
With tricyclic antidepressants (TCAs) there are isolated reports of changes in the INR but the evidence for an interaction
is poor and inconclusive. Monitor INR as normal.
St John’s Wort can decrease the effect of warfarin in some people. The CSM in the UK advise against concurrent use.
There are isolated reports of enhanced anticoagulant effect or hypoglycemia with sulphonylureas but evidence of a clinically
significant interaction with warfarin is very weak and appears to be very unlikely.
The metabolism of warfarin can be increased by carbamazepine, phenytoin and phenobarbitone (reduced anticoagulant effect).
Enhanced anticoagulant effect has been reported with phenytoin and sodium valproate. The evidence for clinically significant
drug interactions is strongest with carbamazepine and phenobarbitone (reduced anticoagulant effect) but consider monitoring
all patients started on concurrent therapy.
Fluconazole inhibits the metabolism of warfarin and a clinically significant interaction occurs in most patients. A
reduction in the dose of warfarin (20–70%) is usually required. Miconazole also inhibits the metabolism of warfarin and
potentially serious bleeds have been reported when miconazole preparations (including oral gel and vaginal preparations)
have been given in combination with warfarin. Miconazole or fluconazole should not be co-administered with warfarin unless
the INR can be very closely monitored. Apart from isolated case reports, there is very little evidence that itraconazole
or ketoconazole interact in the same way.
The anticoagulant effect of warfarin can be enhanced or reduced by ritonavir and nevirapine.
The anticoagulant effect of warfarin is enhanced due to antiplatelet action.
The anticoagulant effect of warfarin may be enhanced or reduced by low to moderate oral doses of corticosteroids (e.g.
prednisone, methylprednisolone, dexamethasone), although the supporting evidence is weak and a clinically significant
interaction does not usually occur. High doses of corticosteroids appear to be more likely to cause a significant effect
Anticoagulant effect possibly enhanced by etoposide, fluorouracil and ifosfamide and reduced by azathioprine and mercaptopurine.
Increased bleeding risk due to antiplatelets effects.
Disulfiram enhances the anticoagulant effects of warfarin and an interaction (increased INR) probably occurs in most
These are often high in vitamin K which may reduce the effect of warfarin.
Foods and Juices
A change in diet, especially changing intake of vegetables, fruits and juices high in vitamin K can affect anticoagulant
control. The range of possible foods that interact ranges from avocados to ice cream. Generally only significant or sustained
dietary changes are likely to have an effect so people should be advised to discuss planned dietary changes with their
GP or pharmacist. Patient information leaflets provide further advice.
Cranberry juice seems to have a different effect and compounds in the juice may inhibit the metabolism of warfarin.
There have been several reports of bleeding and although information is limited the Committee on Safety of Medicines (CSM)
in the UK advise against drinking cranberry juice whilst taking warfarin.
Herbals and supplements
All people taking warfarin should check with their GP or pharmacist before taking herbal preparations or dietary supplements.
Fenugreek, dong quai, ginkgo biloba and garlic are amongst those reported to increase bleeding risk, without a change
in INR (i.e. due to antiplatelet effect). Ginseng has been reported to decrease the INR.
St John’s Wort can decrease the effect of warfarin in some people. The CSM in the UK advises against concurrent use.
Anticoagulant effect can be enhanced by danazol (inhibits metabolism of warfarin), flutamide and tamoxifen. All these
interactions are established and clinically important and concurrent therapy requires close monitoring of the INR. Tamoxifen
and warfarin is the commonest combination seen in practice and dose reductions of warfarin of 30–50 % or more are sometimes
Anticoagulant effect of warfarin possibly enhanced. In case reports an increased INR has occurred after just two or
Lipid lowering drugs
Simvastatin can cause small increases in the INR in some people taking warfarin but a dose reduction of the latter is
not usually required. Atorvastatin does not appear to interact with warfarin. The interaction with fluvastatin and lovastatin
may be more significant.
Cholestyramine has been reported to enhance or reduce the effect of warfarin. Bezafibrate has been occasionally reported
to increase the INR in people taking warfarin.
Oestrogens and progestogens
Anticoagulant effect is antagonised
Pharmacokinetic studies indicate no interaction but fluctuations in anticoagulant control may occur in practice. Consider
monitoring the INR if orlistat is added to warfarin treatment.
There is an increased risk of bleeding when sibutramine given with warfarin.
Thyroid Hormones (thyroxine)
A well documented and clinically important interaction. In people with hypothyroidism who are also taking warfarin,
the addition of thyroxine usually results in an increased anticoagulant effect and a reduction in the dose of warfarin
is necessary. The change in anticoagulant effect may be gradual as the dose of thyroxine is titrated. The addition of
an antithyroid drug (e.g. carbimazole) may necessitate an increase in the warfarin dose.
Vitamin K containing preparation can antagonise the effects of warfarin. Check the vitamin K content of multivitamin
preparations and dietary supplements.
This information is based on;
Stockley’s Drug Interactions. 7th Edition, 2006. Pharmaceutical Press
British National Formulary (BNF). BMJ Publishing Group and Royal Pharmaceutical Society of GB
Bachmann KA (Ed). Drug Interactions Handbook. Lexicomp, Ohio.