Use of INR calibrator plasmas in the routine coagulation laboratory: A study of two thrombolastin reagents

INR values may be either calculated with the ISI values supplied by thromboplastin manufacturers or are directly extrapolated from certified INR calibrator plasmas. We tested the principle of local INR calibration using INR calibrator plasmas (PT-Multi Calibrator, Siemens), two thromboplastin reagents (Neoplastin Plus, rabbit brain, Stago, coagulometer-specific ISI 1.31, and Innovin, recombinant human tissue factor, Siemens) and the same coagulometer (STA-R, Stago) in 100 patients on warfarin. Using a ISI value of 0.77 with Tomenson correction for Innovin (correction factor = 1.09), INR values of patients were similar with the two reagents, with a bias of 0.03 INR units and no significant regression of the difference over the average INR by method comparison analysis. With the INR calibrator plasmas, INR values with Neoplastin Plus were lower than Innovin values with an average bias of 0.39 INR units and a significant regression of the difference over the average INR (r = - 0.91). Significant bias (0.16 INR units, p <0.00001) and regression (r = - 0.77) was also observed by comparison of Neoplastin Plus INRs with Innovin calibrated INRs. Based on a therapeutic INR interval of 2.0 to 3.5, discordance in warfarin dosing was approximately 3 times higher with INR calibration (27% vs 11%). Because of non commutability with fresh plasma samples, local INR calibration with lyophilized calibrator plasmas may not be valid for some reagent-instrument combinations.