Significance of local international sensitivity index systems for monitoring warfarin and liver function

The international normalized ratio (INR): What reagent, what instrument? The assessment of the agreement between INR values according to different reagent/instrument combinations

WHAT IS KNOWN AND OBJECTIVE

The international normalized ratio (INR) is widely used to monitor patients on vitamin K antagonists. This study aimed to assess the agreement of INR values obtained with different thromboplastin/instrument combinations.

MATERIAL AND METHODS

International normalized ratio was determined on plasmas from 330 patients undergoing antivitamin K treatment (with acenocoumarol), using two calibration methods and four reagent/instrument combinations: Both Neoplastine CI and Neoplastine CI Plus on STA-R instrument from Diagnostica STAGO, Asnières, France; and both Thromborel S and Innovin on SYSMEX 2100i instrument from Siemens Health Care Diagnostics, Marbung, Germany. The agreement analysis was done using the Bland-Altman plot and the Cohen Kappa coefficient.

RESULTS

The mean of the differences between the INR values and the limits of agreement were -0.07 [-0.51 to 0.38] for the Neoplastine CI plus and Neoplastine CI reagents, -0.08 [-1.18 to 1.03] for the Thromborel S and Innovin reagents when the INR was calculated, -0.1 [-1.15 to 0.95] for the Thromborel S and Innovin reagents when the INR was directly calibrated and -0.1 [-0.7 to 0.5] for the Neoplastine CI plus and Thromborel S. Cohen's kappa coefficients were 0.94, 0.76, 0.85 and 0.82, respectively.

NEW FINDINGS AND CONCLUSION

The agreement between the four reagent/instrument combinations was high enough to classify patients as inefficaciously or efficaciously anticoagulated. The data interpretation should always be related to the clinical purpose.