Objective: To determine whether computer-aided dosing of warfarin is superior to physician dosing to maintain a patient in a rehabilitation hospital within a target international normalized ratio goal.
Design: Randomized, double-blinded, clinical trial in an inpatient rehabilitation hospital. A total of 30 consecutive patients admitted receiving warfarin were randomized to either clinician dosing or computer-aided warfarin dosing for the duration of their hospitalization. The main outcome measures included the percentage of days in a therapeutic anticoagulation range and the number of blood draws. Exclusion criteria included short length of stay (n=110, 39%) and a physician declared international normalized ratio target range of <2.0 (n=67, 23%). A total of 73 patients were excluded because of heme-positive stools at admission, recent gastrointestinal bleed, early discharge or consent refusal. Dawn AC software was used to determine warfarin dosage and frequency of blood draws to maintain a target international normalized ratio of 2.0-3.0 for the computer-dosed group (n=14). Several physicians recommended warfarin dosages for the second group (n=16). Two were dropped from the computer model secondary to lost data files for these two patients.
Results: Computer-aided dosing of warfarin resulted in 61.7% of days within the therapeutic range (international normalized ratio, 2-3), whereas clinician dosing resulted in only 44.1%. There were no significant differences in the number of blood draws or demographic variables between the two groups.
Conclusion: Computers were significantly better at maintaining patients within a therapeutic international normalized ratio range than physicians. There were no significant differences in the number of recommended blood draws.