FAILURE TO GENERATE COMPARABLE INTERNATIONAL NORMALIZED RATIO VALUES USING 5 DIFFERENT THROMBOPLASTIN REAGENTS IN PARALLEL STUDIES OF PATIENTS RECEIVING WARFARIN

Warfarin therapy has traditionally been monitored using the prothrombi n time (PT). A significant problem with this assay is the variable sen sitivity of commercially available thromboplastin reagents to reductio n of vitamin K-dependent coagulation factors. The International Normal ized Ratio (INR) was developed as a means of standardizing PT values b etween laboratories that use different thromboplastins and types of in strumentation. Parallel testing of samples from 63 patients stabilized on oral anticoagulation with five different thromboplastins was under taken. Forty-eight percent of all samples had INR values that were not identical but showed good correlation. Fifty-two percent of the sampl es had clinically significant discrepancies of their INR values. In th is group, patients in the therapeutic range with one thromboplastin ap peared over- or underanticoagulated based on the INR using a different thromboplastin. In order to determine whether use of a low-ISI thromb oplastin reagent could provide more reproducible INR results, concurre nt testing of specimens from 36 patients on stable oral anticoagulatio n was undertaken between our hospital laboratory and another, nonaffil iated institution. Both hospitals used similar instrumentation and thr omboplastin reagents. Under these conditions, the INR values generated between laboratories correlated highly. The inability to generate con sistent INR values in parallel testing using different thromboplastins with identical samples and instrumentation raises concern about the r eliability of the INR, and suggests that further analysis is necessary to identify the source of these discrepancies. In the interim, our da ta show use of a recombinant thromboplastin with a low-ISI value subst antially improves interlaboratory variation in INR value.