Background: Warfarin, although highly effective in preventing thromboembolism, can cause hemorrhage. For warfarin-treated patients, the risk of bleeding increases as the International Normalized Ratio (INR) rises, particularly if the INR exceeds 4. Although low-dose oral vitamin K reliably decreases supratherapeutic INR values, previous surveys indicate that many providers choose not to administer vitamin K to asymptomatic patients who present with excessive INR prolongation. In effort to help guide the management of asymptomatic outpatients with elevated INRs, we assessed the rate of hemorrhage as well as the frequency of vitamin K use in an observational study.
Methods: A prospective cohort of patients taking warfarin was assembled from 101 sites with personnel designated as warfarin managers across the United States during the period August 2000 to December 2001; 98 sites were community-based physician office practices. Enrolled practices participated in mandatory on-site study training and all patients provided written informed consent. Patients with a first observed episode of INR ≥5.0 were identified. For the two-week period following the index INR, provider-generated outpatient progress notes that pertained to anticoagulation (and were part of the medical record) were individually reviewed by one of the investigators. Outcomes included: site-specific management of elevated INR (recorded vitamin K use versus simple withholding of warfarin), and number of patients sustaining a major hemorrhage within the 2 weeks following the index INR value. Major hemorrhage was defined as fatal, necessitating hospitalization with transfusion of at least 2 units of packed red blood cells, or occurring at a critical site (e.g., intracranial, retroperitoneal).
Results: During the study period, 6,792 patients were enrolled providing 5,961 person-years of follow-up. Of these, 979 had a first observed episode of an INR of more than 5. Thirty-nine % of these patients were receiving warfarin for atrial fibrillation and 29% had a prosthetic heart valve. Mean age was 69 years (range 20–94) and 50% were women. Ninety-six % (n=937) of the INRs were between 5 and 9. Ninety-eight % (n=963) were managed exclusively as outpatients; management could not be assessed in 16 patients who were hospitalized for unrelated reasons.
Patients with an INR >9 were more likely to receive vitamin K compared to those with an INR of 5.0 to 9.0 (62% versus 7%). The mean INR for those patients known to have received vitamin K was 9.9 versus 6.2 among those for whom no vitamin K use was recorded.
Of the 979 patients, 99.7% (n=976) had 14-day follow-up information. Nine patients sustained a major hemorrhage within the 2-week period (0.9%), mean INR 9.4, none were fatal.
Conclusions: Among asymptomatic patients with warfarin-induced coagulopathy, an intervention known to reduce the INR (low-dose oral vitamin K) is used infrequently, particularly for patients whose INR value is 9 or less. For asymptomatic warfarin-treated individuals who are managed in the outpatient setting and whose INR is less than 9, the short-term risk of major bleeding appears to be low. Prospective, controlled trials are needed to determine the risks and benefits of administering vitamin K to this population.