PURPOSE: The objective of the study was to compare in anticoagulated patients the international normalized ratio (INR) measured with a new capillary whole-blood device, the i-STAT Portable Clinical Analyser, with conventional plasma INR obtained from the central laboratory. PATIENTS AND METHODS: Between-cartridge variability was first determined with two lyophilized controls with INR levels of 1.60 and 2.75 (n=10). Next, in 35 patients under different intensities of oral anticoagulation, capillary blood INR was measured with two i-STAT devices and was compared to central laboratory plasma INR (Innovin reagent and BCS analyser). RESULTS: Between-cartridge coefficients of variation were 5% (95%, CI 3.4-9.1) and 3% (95%, CI 2.1-5.5) at INR levels of 1.60 and 2.75. Mean INR difference between the two i-STAT devices was 0.1, and the correlation coefficient was 0.98. Between i-STAT and central laboratory INR, the correlation coefficient was 0.95. Bias values were 0.04, 0.2, and -0.04 at INR levels of 2.0, 2.5, and 3.5, respectively. CONCLUSION: The INR measured with the i-STAT Portable Clinical Analyser is precise and compares well with plasma INR performed in a central laboratory.
PURPOSE: The objective of the study was to compare in anticoagulated patients the international normalized ratio (INR) measured with a new capillary whole-blood device, the i-STAT Portable Clinical Analyser, with conventional plasma INR obtained from the central laboratory. PATIENTS AND METHODS: Between-cartridge variability was first determined with two lyophilized controls with INR levels of 1.60 and 2.75 (n=10). Next, in 35 patients under different intensities of oral anticoagulation, capillary blood INR was measured with two i-STAT devices and was compared to central laboratory plasma INR (Innovin reagent and BCS analyser). RESULTS: Between-cartridge coefficients of variation were 5% (95%, CI 3.4-9.1) and 3% (95%, CI 2.1-5.5) at INR levels of 1.60 and 2.75. Mean INR difference between the two i-STAT devices was 0.1, and the correlation coefficient was 0.98. Between i-STAT and central laboratory INR, the correlation coefficient was 0.95. Bias values were 0.04, 0.2, and -0.04 at INR levels of 2.0, 2.5, and 3.5, respectively. CONCLUSION: The INR measured with the i-STAT Portable Clinical Analyser is precise and compares well with plasma INR performed in a central laboratory.