Objective: To measure the rate of decrease of the international normalized ratio (INR) after temporary discontinuation of warfarin therapy.
Design: Prospective evaluation of an outpatient cohort.
Setting: University medical center anticoagulation clinic.
Patients: 22 patients receiving a fixed evening dose of warfarin for whom temporary discontinuation of therapy was deemed safe.
Measurements: Serial plasma samples were drawn for INR measurements approximately 20, 65, 115, and 185 hours after patients received the last dose of warfarin. In five patients, INR was measured twice daily for 5 days.
Results: For patients with a mean steady-state INR of 2.6, the mean INR 65 hours (2.7 days) after discontinuation of warfarin therapy was 1.6 (range, 1.11 to 2.16); 20 of 22 patients (91%) had an INR greater than 1.2. The mean INR 115 hours (4.7 days) after discontinuation of warfarin therapy was 1.1; 5 of 22 patients (23%) had an INR of 1.2 or greater. In 5 patients studied in detail, the INR decreased exponentially and had a half-life that ranged from 0.52 to 1.2 days; the onset of maximal decrease began 24 to 36 hours after discontinuation of warfarin therapy. In the total cohort, age was a significant (P < 0.005) independent predictor of smaller decreases in the INR between day 1 and day 3 (regression coefficient = −6.8% ± 2%/2 days per decade of age; R2 = 0.34).
Conclusions: By simulating preoperative discontinuation of warfarin therapy, we found that the INR decreases exponentially, with wide interpatient variation in the rate of decrease. Age is associated with a slower rate of decrease. To be certain that the INR at the time of the surgery is less than 1.2, warfarin should be withheld for 96 to 115 hours (4 doses) in patients with a steady-state INR between 2.0 and 3.0. For patients with a higher steady-state INR, a longer wait is necessary.