A comparison of INRs after local calibration of thromboplastin international sensitivity indexes.

There are approximately 300 reagent/instrument combinations for performing prothrombin times/international normalized ratios (PT/INR) in the United States. Manufacturers and laboratories continually struggle to ensure that the International Sensitivity Index (ISI) of their thromboplastin is accurate for assaying PT/INR. OBJECTIVE: This study reports the feasibility of a new method to locally calibrate ISI of thromboplastin using the mechanical STA automated coagulation analyzer (Diagnostica-Stago Inc.) and two photo-optic coagulation analyzers, the BCS (Dade-Behring) and CA-540 (Sysmex). DESIGN: Neoplastine CI+ (CI+) (Diagnostica-Stago Inc); Thromboplastin C+ (TC+); Thromborel S (TRS); and Innovin (I) (Dade-Behring) were used in this study. A mean normal PT (MNPT) was determined for each reagent/instrument combination using samples from 25 normal individuals. Manufacturer instrument specific ISI values were not available for the STA with TC+, TRS and I. The CA540 had no ISI value for CI+ and the BCS system had no manufacturer assigned ISI values for TC+ and I; generic photo-optic and mechanical ISI manufacturer values were used for these two systems. Local on-site calibration was performed using frozen plasma calibrators to determine ISI values for each thromboplastin. Post-calibration, 95 patient samples were assayed for each reagent/instrument system combination using the manufacturer ISI and the local calibrated ISI to determine the INR result. PATIENTS: Patients from whom samples were obtained included five with a lupus anticoagulant, 30 on heparin therapy, and 60 on coumadin therapy. RESULTS: Differences between manufacturer versus local calibrated ISI ranged from 0.9% to 18.9% for normal sample INRs and from 0.8% to 16.4% for patient sample INRs. The number (or proportion) of patient specimens with clinically significantly different INR values (>10.0% difference) ranged from zero for several reagent combinations to more than half (or >50.0%) of those tested for several other combinations. CONCLUSION: Our results indicated that by locally calibrating ISI values, each laboratory may eliminate variability and guesswork between different reagent/instrument systems for ISI values when performing PT/INR assays and potentially improve the clinical accuracy of their patients' PT/INR results.