Background/purpose: Warfarin treatment benefits vary with the clinical skill of warfarin dosage adjustment. However, optimal dosage adjustment in response to the out-of-range international normalized ratio (INR) is not completely understood.
Methods: Data for 2014-2016 from an integrated health care information system of a single hospital were analyzed. Based on their warfarin dosage fluctuation (WDF), defined as the standard deviation of all prescribed warfarin dosages divided by the mean dosage, the patients were classified into Groups 1 (0-0.10), 2 (0.10-0.20), and 3 (>0.20). Target and in-range INRs were 2.0-2.5 and 1.5-3.0, respectively. Based on time in therapeutic range (TTR), patients were also categorized into Groups 1 (<0.6), 2 (0.6-0.9), and 3 (>0.9). The primary outcome was emergency department visits for bleeding or thromboembolism (TE) events.
Results: Eighty-three events were retrieved from 17,397 records (1834 patients). Annual incidence rates were 0.9%, 2.5%, and 4.5% for WDF Groups 1, 2, and 3, respectively (p < 0.05), and 2.3%, 1.7%, and 1.7% for TTR Groups 1, 2, and 3, respectively (p = 0.41). Area under the receiver operator characteristic curves for WDF and TTR were 0.686 and 0.519, respectively, indicating that WDF exhibited superior predictive performance than TTR.
Conclusion: High WDF rather than low TTR was associated with increased bleeding and TE incidence rates.
Keywords: Bleeding; International normalized ratio; Thromboembolism; Warfarin; Warfarin dosage fluctuation.
Copyright © 2018. Published by Elsevier B.V.