Correction of excessive anticoagulation with low-dose oral vitamin K1

Authors: Weibert R, Le D, Kayser S, Rapaport S.
Source: Annals of Internal Medicine. 125:959-62. June 15, 1997.
Institution: University of California, San Diego and San Francisco.
Financial support: National Institutes of Health



Patients receiving warfarin anticoagulation occasionally are over-anticoagulated, placing them at risk for bleeding. Although oral vitamin K can reverse the effects of warfarin, it is seldom used in this setting. Because of concern about causing subsequent warfarin resistance, vitamin K is often not administered, while warfarin is temporarily withheld waiting for the INR to decrease. When vitamin K is used, it is usually given parenterally because of concern about the bioavailability of the oral formulation. Studies from the late 1950's showed that oral vitamin K is, in fact, effective in this situation. The authors describe their experience with this approach in managing outpatients with INR's greater than 5.0 and without severe bleeding.


Data was analyzed retrospectively from the records of the anticoagulation clinics of the University of California San Diego (66 patients) and San Francisco (15 patients). All patients had INR's exceeding 5.0 and were treated with oral vitamin K. None of them had severe bleeding, but 7 of the 66 San Diego patients had minor bleeding (hematuria, gingival, hemoptysis and vaginal). 2.5 mg of vitamin K was administered orally to all patients. 58 patients had one dose of warfarin held and returned for a repeat INR measurement the next day; 23 were unable to return the next day, had two doses held and returned after 48 hours. Warfarin was restarted at an adjusted dose when the INR was less than 5.0. A third INR measurement was obtained 4 to 7 days after warfarin was restarted.


The indications for anticoagulation were primarily venous thromboembolism (34 patients), prosthetic heart valves (19) and atrial fibrillation (15). Results are reported separately for patients who returned after 24 and after 48 hours.

Among 76 patients who had a repeat INR 4 to 7 days later, again on warfarin, the results were as follows:

The original mean daily dose of warfarin was 5.2 mg; subsequently reduced to 3.8 mg. There were no instances of clinically evident thromboembolism.

Authors' Discussion

The authors note that there are two objectives in administering vitamin K to over-anticoagulated patients: to bring the INR down to a therapeutic range quickly (to less than 5.0 within 24-48 hours), while not overshooting the mark (INR less than 2.0) and not causing resistance to warfarin.

The first objective was achieved in 90% of patients (96% of patients if those with INR's greater than 10.0 are excluded).

The second objective was achieved in 83% of patients; only 6% of patients had an INR less than 1.8.

They conclude that administering 2.5 mg of vitamin K orally is worth considering in over-anticoagulated patients who do not have evidence of significant bleeding, whose INR is not greater than 10.0 and in whom a transient episode of under-anticoagulation is not deemed to be a major risk. Further studies are recommended to refine the approach, and to compare it to simple withholding of warfarin.


This study fits in well with the trend towards reviving some lower-tech approaches from the past. Oral administration of a small dose of vitamin K may be reasonable in the patients described above. I have a few caveats about this study, however.

Since the INR was measured 24 or 48 hours after the dose of vitamin K, and then again 4 to 7 days later, it is likely that some patients whose INR became lower in the interim and then increased were missed. The risk for being subtherapeutic was probably underestimated by this study.

Since these patients were being followed in an anticoagulation clinic, where the subcutaneous administration of vitamin K would be very easily performed, the advantage of the oral route over the SQ route does not seem particularly great.

Finally, it would have been interesting to know whether the effect of 2.5 mg of vitamin K depended on the actual dose of warfarin that patients were receiving. It seems intuitively reasonable that patients on 10 mg of warfarin would have less of a response to a given dose of vitamin K than those on 2.5 mg, but that may not be the case.

June 29, 1997

Reader comments

Date: Wed, 30 Jul 1997
From: satya kumar <>

Dr Pinjala RK,
Vascular Surgeon, Andhra pradesh, India

I am very happy to know that oral Vit K can be used to correct the excessive oral anticoagulation effect in some of the DVT patients who are on long term oral anticoagulation. In my country where people come from far off rural areas, many are scared to know that they have to take injections in case of excessive oral anticoagulation. I am sure if the other studies can also confirm the findings of this study, many patients would prefer this method. I am sure we would also do a similar study in our hospital. I congratulate the authors for this simple but very useful study which can certainly comfort our patients.

Dr Pinjala Ramakrishna.,
Vascular surgeon,
Nizam's Institute of Medical Sciences,
Hyderabad - 500082

Submit a comment about this article


Search Medline for "anticoagulants/adverse-effects and vitamin K" from the NLM's PubMed database.

Return to Journal Club home page

Site Meter