The reliability of international normalized ratios during short-term oral anticoagulant treatment.

The reliability of the international normalized ratios (INR) system in the induction phase of coumarin administration has been studied in 15 serial patients over the first 7-40 days of treatment (mean 13.1). The INR results obtained with a variety of thromboplastin reagents have been compared with those obtained with the WHO second primary IRP, BCT/253. A wide divergence of INR values was observed with the various thromboplastins on each day of testing. INR values cannot therefore be relied upon with some of these reagents in the early days of anticoagulant treatment. This probably arises from the difference in responses of the thromboplastins to depression of vitamin K-dependent clotting factors. Consistent deviations from the IRP suggested that additional error may be due to inaccurate calibration of their products by the manufacturers. When the slopes of the sensitivity of the individual reagents to clotting factors II, VII and X were compared, however, results overall more closely approximated to those of the IRP when the INR were substituted for simple prothrombin ratios.