Vitamin K to Reverse Excessive Anticoagulation: A Review of the Literature

Charles T. Taylor, Pharm.D., Elizabeth A. Chester, Pharm.D., Debbie C. Byrd, Pharm.D., and Mark A. Stephens, Pharm.D., Department of Pharmacy Practice, Auburn University School of Pharmacy, Tuscaloosa, Alabama; the Departments of Internal Medicine and Family Medicine, University of Alabama, School of Medicine, Tuscaloosa Program, Tuscaloosa, Alabama; the Department of Family Medicine, University of Alabama, School of Medicine, Huntsville Program, Huntsville, Alabama; and the Department of Pharmacy Practice and Pharmaco-economics, The University of Tennessee College of Pharmacy and the Regional Medical Center, Memphis, Tennessee.

Pharmacotherapy. 1999;19(12) 

In This Article

Trials of Vitamin K Administration

Practice guidelines recommend vitamin K to reverse anticoagulation in three situations[1]: for asymptomatic patients with elevated INR, who are considered at increased risk of bleeding, and in whom it is important to reduce the INR to less than 5.0 within 24 hours; for patients who require surgery, in which case it is reasonable to reduce the INR to less than 1.5 at the time of surgery; and for patients with serious bleeding to reduce the INR to 1.0 as soon as possible (Tables 2-4).[8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25] Unless otherwise noted, studies in the tables incorporated temporary warfarin discon-tinuation together with vitamin K administration.

Intravenous administration of vitamin K is safe and effective in reversing the effects of oral anticoagulation.[9,10,11,12] High doses (>= 10 mg) produce rapid reversal; however, overcorrection and warfarin resistance complicate therapy.[9,10,11,12] In a pharmacodynamic evaluation, the effects of warfarin on clotting factor synthesis were antagonized for 7 days by a 10-mg dose of vitamin K.[29] Low doses (0.5-2.5 mg) reversed anticoagulation within 24 hours with less risk of overcorrection or resistance to reinstitution of therapy.[1]

Some adverse effects are specific to intravenous administration and may resemble anaphylaxis or hypersensitivity. Symptoms may include facial flushing, diaphoresis, chest pain, hypotension, and dyspnea.[1,8] Rare case reports documented cardiac arrest and death. Patients were usually elderly, seriously ill, and receiving vitamin K doses of 5-20 mg.[8,25,30] Product information contains a black box warning recommending intravenous administration only when other routes are not feasible and the risk is justified.[31]

Eighteen hospitalized patients with a British ratio (BR) over 5.0 (therapeutic 2.0-3.0) were randomized in a clinical trial to compare the value of prothrombin complex concentrate (9 patients) with intravenous vitamin K 2.5 mg (9 patients) in reversing excessive anticoagulation.[9] The authors did not state whether patients were actively bleeding nor describe specific exclusion criteria. The degree of overanticoagulation in this patient population due to nicoumalone and warfarin overdose warranted rapid reversal [prothrombin time (PT) range 40-280 sec; BR range not reported]. Immediate (within 30 min) and maximum correction of PT, partial thromboplastin time, and clotting factor assays occurred with the concentrate, but these changes were not always sustained. At 24 hours reversal was thought to be insufficient in two (22%) of nine patients (BR 10.0 and 10.5). In contrast, intravenous vitamin K showed little benefit at 2 hours and consistent overcorrection at 24 hours despite relatively conservative dosing. No complications occurred; however, one patient developed hepatitis 8 weeks after receiving prothrombin complex concentrate. The concentrate provided a more rapid and controlled, but less sustained, effect than intravenous vitamin K.

Seven hospitalized patients had significant bleeding and PT elevations greater than 27 seconds (mean PT 39.3 sec, range 27.8-80 sec).[10] Six patients received intravenous vitamin K 1 mg and one received 0.5 mg to reverse anticoagulation to therapeutic levels (PT range 15-33 sec, 1.5-2.5 times control values). Five patients (71%) experienced an adequate response (PT 14.9-21.3 sec) within 8-21 hours. Both patients with inadequate response had received an initial 1-mg dose of the vitamin. One patient received an additional 2 mg, resulting in a PT of 12.4 seconds at 18 hours. Another patient with an initial PT of 80 seconds did not receive additional vitamin K and the PT fell to 28.6 seconds at 75 hours. Bleeding was controlled in all patients without requiring fresh-frozen plasma. The results support administration of low-dose vitamin K.

Thirty-one patients with an INR above 5.0 were included in a study to evaluate the efficacy of low-dose intravenous vitamin K 0.5 and 1 mg in patients not actively bleeding.[11] Ten consecutive patients (mean INR 10.5, range 6.4-17.8) received 1 mg and 21 (mean INR 10.3, range 5.6-25.9) received 0.5 mg. All 10 patients who received vitamin K 1 mg had an INR less than 4.5 at 24 hours, with mean reduction of 75%. However, five of them overcorrected at 24 hours, with INRs less than 2.0; values after 24 hours were not reported. In the 21 patients who received vitamin K 0.5 mg, overcorrection was not documented at 24 hours but 1 patient overcorrected at 48 hours. Within 24 hours, 18 patients (86%) achieved an INR below 4.5, with mean reduction of 62%. No adverse effects were documented. The authors concluded that a low dose of intravenous vitamin K (0.5 mg) is an effective and convenient way to control oral anticoagulation.

A prospective study enrolled 23 consecutive patients with an INR above 10.0.[12] Patients requiring transfusion of blood products for immediate treatment of active bleeds were excluded. The effect of temporary warfarin discontinuation alone versus escalating low doses of intravenous vitamin K was assessed. The first six patients were observed without treatment beyond temporary warfarin discontinuation (group 1). Intravenous vitamin K was administered to all subsequent patients in escalating doses: 0.1 mg to one patient and 0.2 mg to three patients (group 2), 0.5 mg to four (group 3), and 1.0 mg to nine (group 4). Mean INR values were not reported for each group. At approximately 24 hours, only one patient in group 1 had an INR less than the goal of 5.0. In fact, two patients developed signs and symptoms of spontaneous hemorrhage. No patient in group 2 had an INR below 5.0 at 24 hours, and only one in this group had a therapeutic INR (2.0-3.5) within 48 hours. Three of four patients in group 3 achieved an INR below 5.0 by 24 hours, yet only one patient had an INR below 3.5. All nine patients in group 4 had an INR below 3.5 (range 1.7-3.5) within 24 hours of treatment, with only one patient's INR falling below therapeutic range. No patient who received intravenous vitamin K had adverse effects. The authors concluded that intravenous vitamin K 1 mg is the optimal dose to reverse excessively anticoagulated patients with INRs in excess of 10.0.

In 1995 PP recommended subcutaneous vitamin K to reverse anticoagulation in non-emergencies.[7] This route was favored over intravenous administration due to fewer reports of anaphylactoid reactions.[1,7] In 1998 PP removed subcutaneous vitamin K from its recommendations based in part on a retrospective study suggesting that the response may be unpredictable or delayed.[1,25] Four published trials, one retrospective, two prospective, and one prospective, randomized study, evaluated subcutaneous vitamin K for this indication.

A prospective study[13] evaluated applicability of the recommendations of the Third Conference on Antithrombotic Therapy and reliability of subcutaneous vitamin K for warfarin-induced INR elevations.[6] A control group of 6 hospitalized patients with elevated INRs (mean 7.98, range 4.6-18.0) were treated by temporary warfarin discontinuation and compared with 12 patients with elevated INRs (mean 12.57, range 6.0-25.3) managed with subcutaneous vitamin K 1-20 mg. Patients with signs of active bleeding were excluded. Vitamin K dosing was based on previous PP guidelines.[6] A mean dose of 4.9 mg was necessary to return patients to an INR of 3.0 or less. The average time to reach this level of reversal was 31 hours in the treatment group and 49.8 hours in the control group. The 12 patients who received subcutaneous vitamin K achieved an INR of 3.0 or less within 4-50 hours of administration. Two of them (INRs 19.2 and 22.1) experienced oozing from intravenous sites and received vitamin K 20 and 10 mg, respectively.

A retrospective comparative study in a university-affiliated hospital evaluated 148 episodes of asymptomatic excessive anticoag-ulation from acenocoumarol (INR > 6.0) in 68 patients (mean INR 8.7, range not reported).[14] Doses of subcutaneous vitamin K were 0.5 mg for INRs below 10.0 and 2.0 mg for INRs above 10.0. The comparative group consisted of 80 patients (mean INR 7.4, range not reported) treated by discontinuation of oral anticoagulation for approximately 1-2 days. Mean reduction in INR in the first 24 hours was 3.12 (42%) in the discontinuation group and 5.8 (62%) in patients receiving vitamin K (p=0.01). Twenty-four hours after therapy, 94% of treated and 75% of nontreated patients had INRs below 6.0. Overcorrection (INR < 2.0) was noted in the treated (79%) and discontinuation (78%) groups at 48 hours. No thrombotic complications occurred; however, one episode each of minor (epistaxis) and major (gastrointestinal) bleeding was reported in patients receiving vitamin K. Thus low-dose (0.5-2.0 mg) subcutaneous vitamin K plus 24-hour discontinuation of anticoagulation was superior to discontinuation alone for asymptomatic patients.

The efficacy and safety of subcutaneous vitamin K were evaluated in 21 excessively anticoagulated patients (INR range 8.0-17.6) recruited from an anticoagulation clinic and an emergency department.[15] Patients with evidence of bleeding were excluded. Group 1 consisted of 17 patients (INR >= 8.0 but < 14.0, mean 8.9) who received 1 mg of the vitamin, and group 2 consisted of 4 (INR >= 14.0 but < 20.0, mean 15.9) who received 2 mg. At 24 hours, only 47% of group 1 patients achieved an INR below 4.5; however, 93% reached this value at 48 hours (mean reductions 49% and 72% at 24 and 48 hrs, respectively). Only 25% of group 2 patients had an INR below 4.5 at 24 hours, whereas 100% achieved it within 48 hours (mean reductions 67% and 85% at 24 and 48 hrs, respectively). The INR in four patients (23%) in group 1 and one patient (25%) in group 2 fell below 2.0 at 48 hours. No patient experienced hemorrhagic or thrombotic complications.

The only randomized, prospective study was conducted in a tertiary care hospital.[16] Fifty-five excessively anticoagulated patients (INR > 6.0) with no evidence of active bleeding were randomized to receive subcutaneous (33 patients, mean INR 8.4, range 6.0-13.0) or intravenous (22 patients, mean INR 7.9, range 6.0-14.0) vitamin K. For an INR between 6.0 and 10.0, patients received 0.5 mg of the vitamin (25 subcutaneous, 20 intravenous). Patients with an INR above 10.0 received 3.0 mg (8 subcutaneous, 2 intravenous). Follow-up INR values were determined 24 and 72 hours, and 7 days after administration. Significantly more patients (95%) receiving intravenous vitamin K had INRs below 5.0 at 24 hours compared with subcu-taneous administration (45%, p<0.001). At 72 hours, all patients but one had INRs below 5.0. Of note, 5 (23%) of 22 patients receiving intravenous therapy and 14 (42%) of 33 receiving subcutaneous therapy experienced overcorrection (INR < 2.0, p=0.13). No bleeding or thrombo-embolic complications were reported; however, the study had inadequate power to determine a difference in rates of anaphylaxis, bleeding, or thromboembolic complications. Although intravenous therapy resulted in a more prompt reduction in INR, subcutaneous vitamin K was an effective and safe alternative.

Reports of orally administered vitamin K first appeared in the medical literature in the early 1950s when the intravenous formulation given by mouth proved beneficial in reducing excessive anticoagulation. Ease of administration improved in 1955 with introduction of an oral tablet.[17] In the 40 years since then, a handful of case reports and clinical investigations continued to describe the therapeutic utility of oral vitamin K to correct or reverse warfarin's anticoagulant effect.

Experience with the newly available vitamin K tablet was described in 75 excessively anti-coagulated, hospitalized patients who received a single dose of 2.5, 5, 10, 15, or 20 mg.[17] Of 15 patients receiving 2.5 mg (PT 35-49 sec), 73% had adequate reversal of anticoagulation (defined as PT < 35 sec) within 12 hours. All but one of these patients had a PT less than 35 seconds at 24 hours. Two patients (13%) in this group had anticoagulation overcorrected below the therapeutic range (PT < 22 sec at 12 hrs) and were given heparin. Of 31 patients receiving 5 mg (PT 36-59 sec), 90% experienced adequate reversal of anticoagulation within 12 hours; within 24 hours, 100% responded. Anticoagu-lation status became subtherapeutic in six patients (19%). Of 18 patients receiving 10 mg (PT 38-75 sec), 94% had a therapeutic PT at 12 hours. Only one patient failed to respond within 24 hours, and three (17%) had a PT below the therapeutic range. Of 10 patients receiving vitamin K 15 mg (PT 32-65 sec), 70% experienced adequate reversal and two (20%) overcorrected at 12 hours. One patient (PT 71 sec) received 20 mg and 12 hours later his PT was subtherapeutic (21 sec). Overall, 85% and 98% of patients had adequate reversal of excessive anticoagulation within 12 hours and 24 hours of receiving oral vitamin K, respectively. In doses of 2.5-15 mg, oral vitamin K was effective and easy to administer and did not produce resistance to continued anticoagulant therapy.

Low-dose oral vitamin K and temporary warfarin discontinuation to correct excessive anticoagulation were compared in 23 hospitalized patients with elevated INRs (> 5.0) and no evidence of bleeding.[18] Patients were assigned randomly to one of two groups: 12 in group A (mean INR 6.14, range 5.11-7.90) discontinued warfarin for 1 day, and 11 in group B (mean INR 5.82, range 5.26-7.5) received oral vitamin K 2.0 mg in addition to their usual warfarin dosage. On day 1 (24 hrs after randomization), only 58% of patients in group A achieved an INR less than 5.0 compared with 100% in group B. By day 2, all but one patient in group A had an INR less than 5.0, and one in group B experienced a rise in INR to 5.3. No patients receiving vitamin K experienced refractoriness to anticoagulation at day 2 or 9.

A retrospective chart review conducted in two university-affiliated anticoagulation clinics assessed management of excessive anticoagu-lation with oral vitamin K in 81 patients with INRs above 5.0 and no evidence of major bleeding.[19] Intervention consisted of withholding one or two doses of warfarin, administering oral vitamin K 2.5 mg, measuring the INR after 24-48 hours, and adjusting the warfarin dosage. In 68 (96%) of 71 patients with an INR between 5.0 and 10.0, vitamin K lowered INR to less than 5.0 within 24-48 hours without inducing resistance to further anticoagulation. However, only 50% of patients (5/10) whose initial INR exceeded 10.0 achieved an INR less than 5.0 within that time. Of 14 patients (17%) who had a subtherapeutic INR at 24 hours, only 5 (6%) had an INR below 1.8 and none had an INR below 1.5.

The thromboembolism units of two university hospitals conducted a prospective cohort study to determine if oral vitamin K 1 mg effectively reduces asymptomatic elevations in INR.[20] Since a tablet formulation was not available, patients received the parenteral product, which was drawn into an insulin syringe and administered by mouth. Asymptomatic patients (49 inpatients, 13 outpatients; mean INR 5.79, range 4.5-9.5) were given the vitamin and instructed to discontinue warfarin therapy until further notice. The mean INR on day 1 (within 16 hrs of vitamin K administration) was 2.86 (range 1.3-8.9). Warfarin therapy was resumed on day 1 in 37 patients (60%). Therapy reliably reduced mean INR in patients receiving warfarin with asymptomatic overanticoagulation (INR 4.5-10.0) and should be considered for all such patients when rapid reversal is not necessary.

A case-control study compared low-dose oral vitamin K and warfarin discontinuation in management of asymptomatic patients with supratherapeutic INRs (5.0-10.0).[21] Fourteen historical controls were used in the warfarin discontinuation arm (mean INR 6.3, range 5.0-8.9) and matched according to initial and target INRs to 14 patients receiving vitamin K 2.5 mg (mean INR 6.34, range 5.08-9.96). Warfarin was resumed in both groups after a therapeutic INR was achieved. Vitamin K significantly reduced the time required to reach a therapeutic INR by 0.9 days compared with warfarin discontinuation (1.4 vs 2.3 days, respectively, p=0.001). Mean reductions in INR after 1 day were 55% and 21% for vitamin K and warfarin discontinuation groups, respectively (p<0.001). Of patients receiving vitamin K, 71% achieved a therapeutic INR by day 1, compared with only 7% who discontinued warfarin. No adverse events occurred, and overcorrection was not noted for patients treated with vitamin K.

A prospective trial studied the efficacy and safety of partially correcting anticoagulation by adjusted-dose oral vitamin K.[22] Subjects were patients anticipating minor surgical or dental procedures (group 1, 30 patients; mean INR 2.6, range 1.7-3.4) and those who were excessively anticoagulated without evidence of bleeding (group 2, 21 patients; mean INR 8.4, range 5.1-13.3). All patients received a single dose of vitamin K (group 1, mean 5.0 mg; group 2, mean 10.0 mg). Patients in group 1 took the dose 36 hours before the procedure and were instructed to continue their daily dose of warfarin. Patients in group 2 received one dose of vitamin K and were instructed to hold their next warfarin dose and resume therapy the next day. The dose of oral vitamin K was calculated using a regression formula {16 - [17 x (INR desired/INR initial)]} and was intended to decrease the INR to a predetermined value. Predicted change in INR correlated with actual change (r = 0.92, p<0.0001). The desired INR (1.5-2.0) was achieved in 97% of group 1 patients before the procedure and in all group 2 patients within 48 hours.

A prospective study examined the efficacy of oral vitamin K in 24 patients with INRs above 6.0 (mean 9.3, range 7.3-14.2) receiving phenpro-coumon, a coumarin derivative.[23] Patients temporarily discontinued phenprocoumon and received a mean dose of the vitamin 2.7 mg (range 1-5 mg) based on initial INR values and target zone of anticoagulation according to an algorithm employed at the clinic. For a target zone of anticoagulation in the range of 2.0-3.5, the dose of vitamin K was 2, 3, or 5 mg for initial INR values of 8.0-8.9, 9.0-11.9, and 12.0-14.9, respectively. It was optional to give vitamin K for INRs between 6.0 and 7.9. The proportional decrease in mean INR was 40% in 24 hours, with a mean INR of 4.3 in 48 hours. Sixteen patients (67%) achieved INRs below 6.0 at 24 hours, with two patients (8%) overcorrecting on days 2, 3, and 4. After 1 week only seven patients (29%) remained within the therapeutic INR range (2.0-4.0); the rest were supratherapeutic after an initial reduction. After day 2 the mean INR did not decrease further, but increased approximately 2-8%/day. No episodes of bleeding or thrombo-embolic complications were noted. The authors concluded that treatment of overanticoagulation should be intensified to improve its efficacy.

Two retrospective reviews compared different routes of vitamin K administration in manage-ment of excessively anticoagulated patients. One compared warfarin discontinuation with vitamin K given by subcutaneous, intravenous, intra-muscular, or oral route.[24] Initial INRs were 6.0 or above, and patients were treated according to physicians' judgment. The retrospective analysis included 301 episodes of excessive anticoagu-lation in 248 patients (mean INR not reported, range 6.1-41.8). Episodes were categorized into INRs of 6.0-10.0 (234 patients), 10.1-20.0 (50), and greater than 20.0 (17). Most episodes (83%) were managed by temporary warfarin discontin-uation until the INR fell to within the therapeutic range. Warfarin was discontinued for an average of 3.2 days, and the weekly dose was decreased by approximately 25% from baseline when therapy was resumed. Two major bleeding events (0.8%; initial INRs 7.0 and 6.6) and 35 minor bleeding events (13%) occurred with warfarin discontinuation. In contrast, 16 major bleeding events (31%; initial INRs not reported) and 10 minor bleeding events (20%) occurred in the vitamin K group. These patients were bleeding before vitamin K administration, and 12/16 major bleeds occurred at INRs above 10.0. Two patients developed thromboembolism after receiving vitamin K 10 mg (orally) and 50 mg (20 mg orally, 30 mg subcutaneously), respectively. Three patients became refractory to warfarin and required heparin therapy. Of episodes treated with vitamin K, a mean dose of 10.8 mg was given subcutaneously (39%), intramuscularly (38%), orally (16%), and intravenously (7%). Based on these results, administration of vitamin K to patients with moderate INR elevations (6.0-10.0) may be unnecessary, and warfarin discontinuation is a reasonable alternative.

The other review evaluated the effectiveness of different doses (based on physicians' judgment) of the vitamin 0.5-10 mg intravenously, 1-10 mg subcutaneously, and 2.5 or 5 mg orally to reverse excessive anticoagulation.[25] The 32 patients with elevated INRs (mean 12.5, range 5.7-37.8) were categorized into four groups: low-dose intravenous (LDIV) 0.5 mg or less, high-dose intravenous (HDIV) 1-10 mg, subcutaneous, and oral therapy. On average, eight patients in the LDIV group (mean INR 11.9, range 6.8-17.0) received vitamin K 0.4 mg; this reduced mean INR to below 5.0 within 48 hours in six patients (75%), with one experiencing overcorrection. Nine patients in the HDIV group (mean INR 13.9, range 6.6-21.5) received on average 4.2 mg, which reduced mean INR to below 5.0 within 24 to 48 hours in all patients, with four experiencing overcorrection. In 10 patients (mean INR 14.9, range 5.7-37.8) who received subcutaneous vitamin K (mean dose 2.5 mg), mean INR was reduced to below 5.0 within 24-48 hours in 7 (70%), with no overcorrection. In five (83%) of six patients (mean INR 9.4, range 7.8-13.1) oral vitamin K (mean dose 3.8 mg) reduced mean INR to below 5.0 within 24-48 hours, with no overcorrection. Although HDIV was most effective in lowering INR to below 5.0, it was associated with greatest overcorrection. Failure to achieve an INR below 5.0 was a greater problem in the subcutaneous group (3 subcutaneous, 2 LDIV, 1 oral); however, the initial mean INR was 20.0 for these patients.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: