Inaccuracy of point-of-care international normalized ratio in rivaroxaban-treated patients

To report 2 cases in which point-of-care international normalized ratios (POC-INRs) measured using a Hemochron Jr. Signature Elite device (International Technidyne Corporation) were inaccurate in rivaroxaban-treated patients.

Therapy in an 86-year-old man with atrial fibrillation was converted from warfarin to rivaroxaban 15 mg twice daily because of a deep venous thrombotic event despite an INR of 2.4, which was within the therapeutic range. One week later a POC-INR was inadvertently measured, which was 6.3. In light of the POC-INR being markedly elevated, a laboratory test for INR was performed, which gave a result of 2.74. Therapy in a 66-year-old man was converted from war-farin to rivaroxaban 15 mg twice daily because of unstable INRs and a pulmonary embolism despite a therapeutic INR. Seven days after rivaroxaban was started, the patient's POC-INR was 9.2; simultaneously measured laboratory-determined INR was 2.0. For both patients, coagulation tests performed on follow-up visits revealed continued discordance between the POC and laboratory assays.

Rivaroxaban is an oral factor Xa inhibitor with a predictable pharmacokinetic profile, allowing for a fixed-dose regimen without the need for coagulation monitoring. When patients' therapy is switched from rivaroxaban to warfarin, it is recommended that the drugs be given concurrently until the INR is 2.0 or higher, to ensure adequate anticoagulation during this well-recognized vulnerable period for stroke. POC testing is a common method of INR assessment in clinical practice. During ROCKET-AF (Rivaroxaban Once-Daily, Oral, Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation), POC-INRs were measured exclusively with the INRatio device (Hemosense), and values above 4 were seen very rarely (0.25%), which indicates that the values determined in our patients were highly unusual.

Our 2 patients receiving rivaroxaban had POC-INRs elevated beyond what was expected; these measurements were discordant from INRs simultaneously measured via the laboratory. A prospective evaluation assessing the accuracy of other commonly used POC-INR devices in patients receiving rivaroxaban would determine whether our findings extend to other devices. Until that time, laboratory measurement of INR or POC-INR using an INRatio device is recommended when patients' therapy is transitioned from rivaroxaban to warfarin.