ABSTRACT
BACKGROUND: Initiating warfarin in patients with atrial fibrillation at a typical loading dose of 10 mg daily for 2 days may be associated with excess anticoagulation, especially in older patients. In addition, daily monitoring associated with this regimen may be inconvenient for outpatients. Initiating warfarin at 5 mg daily may produce more consistent anticoagulation and eliminate the need for daily monitoring. The investigators tested whether administering 5 mg warfarin daily for 4 days and checking the anticoagulation status via the international normalized ratio (INR) on the fifth day could predict an early warfarin maintenance dose.
POPULATION STUDIED: This study included 61 outpatients with nonrheumatic atrial fibrillation, not receiving heparin, between the ages of 42 to 88 years (mean age = 71 years). Patients were excluded from the study if they were being treated with a drug known to interact with warfarin or had a coagulation disorder, contraindication to warfarin therapy, previous course of anticoagulation treatment, or baseline INR > 1.2, or if they refused to participate.
STUDY DESIGN AND VALIDITY: This was a prospective cohort study in which patients were given 5 mg warfarin daily for 4 days (day 1 to day 4). The INR was measured on day 5 and the patient’s physician freely chose a new dose. The INR was checked at least once a week for the next 2 weeks. Patients were followed for 3 months. The warfarin dose was considered stable when the INR was between 2 and 3 on 3 consecutive occasions at least 1 week apart. The stable weekly warfarin dose was plotted against the INR obtained on the fifth day of initiation to establish a scheme for predicting an early warfarin maintenance dose. To test the validity of this scheme, 23 additional patients with nonrheumatic atrial fibrillation were given the predicted warfarin maintenance dose based on their INR on day 5 and were followed for 3 months. Although the characteristics of this additional group of patients were not stated, these subjects were enrolled according to the same criteria.
OUTCOMES MEASURED: The outcome of this study was to determine whether the INR on day 5 of warfarin treatment could predict an early warfarin maintenance dose. Major and minor bleeding episodes and thromboembolic complications were also measured.
RESULTS: Of 91 patients eligible to participate in the study, 61 were included in the analysis. Thirty patients could not be evaluated either because they failed to reach a therapeutic INR within 3 months or because they did not complete the follow-up period. The relationship between the weekly maintenance dose and the INR on day 5 followed a hyperbolic curve, demonstrating a direct relationship between the INR on day 5 and the weekly warfarin dose. During the evaluation phase that was conducted in 23 additional patients, the mean difference between the predicted and actual doses of warfarin was 1.6 mg/week (95% CI, .0007-3.195 mg) with a maximum difference of 9 mg/week. One minor bleeding episode was reported among the 61 initial patients and 23 additional patients. The dosing regimen, based on the INR after 4 days of 5 mg warfarin, is shown in the Table.
TABLE
WARFARIN DOSING REGIMEN USED IN STUDY
| INR on Day 5 | Approximate Daily Warfarin Dose (mg)* |
| 1.3 | 6 |
| 1.4 | 5 to 6 |
| 1.5 | 5 |
| 1.6 | 4 to 5 |
| 1.7 | 4 to 4.5 |
| 1.8 | 4 |
| 1.9 | 3.5 to 4 |
| 2.0 | 3.5 |
| *Doses are rounded and calculated from a cumulative weekly dose. | |
Starting outpatients with 5 mg per day of warfarin and basing a maintenance dose on the INR obtained on the fifth day is an effective way to initiate therapy. The difference between the actual and predicted maintenance doses was small, indicating that this simple scheme is a good model for predicting the warfarin maintenance dose. Although only 23 patients were tested using this warfarin dosage scheme, clinicians may consult it when choosing a warfarin maintenance dose rather than using the trial-and-error method that is often pursued in daily practice.