Abstract P199: Randomized Comparison of the DAWN AC Computer Program and a Simple Manual Nomogram for Quality of Warfarin Dosing

Abstract

Background In patients receiving warfarin, the quality of anticoagulant control as measured by the time-in-therapeutic range (TTR) for the International Normalized Ratio (INR) is a key determinant of risk for thromboembolic and bleeding events. Computer programs can assist physicians in optimizing TTR, but are expensive. The purpose of this study was to determine whether a computer system (DAWN AC) was non-inferior to a two-step manual nomogram used in a hospital anticoagulation clinic. Methods Stable anticoagulated patients receiving warfarin with target INR range 2-3 in the anticoagulation clinic were randomized to management with the newly acquired DAWN AC computer system or the clinic's standard of care, a simple manual dosing nomogram. After an initial run-in phase, study data collection started on February 1 st 2010 and was completed on August 8 th 2010. Primary outcome was the mean TTR calculated by the Rosendaal linear interpolation method. The non-inferiority margin was set at 4.5% lower TTR for DAWN AC compared with the nomogram. Results Of the 1,298 patients initially randomized, 1,127 were still managed by the clinic after the run-in phase and entered the study on February 1 st ; 564 were managed with DAWN AC, and 563 with the manual nomogram. The mean age of study patients was 69 ± 14 years and 62% were male. Main indications for anticoagulation were atrial fibrillation (48%) and prosthetic heart valves (25%). Mean follow-up was 172 days, encompassing 8,344 INR values and 155,041 patient days. Adherence to recommended warfarin doses was higher in the DAWN AC than in the nomogram group (99 vs. 90%; p<0.0001), the average interval between INR measurements was similar in the two groups (21 ± 12 vs. 21 ± 13 days; p=0.1987). In the primary analysis, mean TTR in the DAWN AC group was non-inferior to mean TTR in the nomogram group (71.0% ± 23.3 vs. 71.9% ± 22.9; non-inferiority p=0.0052). Conclusion Among stable patients receiving warfarin with a target INR of 2-3 in an anticoagulation clinic, quality of anticoagulant control with the DAWN AC computer program was non-inferior to a simple two-step manual dosing nomogram. The nomogram could be a useful dosing tool for physicians without access to a computerized warfarin dosing system.