Oral anticoagulant therapy control: evidence that INR expression improves the inter-laboratory comparability of results--the Bologna oral anticoagulant control exercise.

The aims of the present study were: 1) interlaboratory normalization of prothrombin time (PT) testing for anticoagulant therapy control through calibration of customary thromboplastins against international reference materials, and 2) "on field" validation of the advantages offered by expression of results as International Normalized Ratio (INR) as opposed to percentage activity. PT tests were carried out over 8 days on the same normal subjects (16) and patients on oral anticoagulants (48) in the 9 laboratories of the Bologna area. The use of customary thromboplastins and coagulometers was maintained in all labs throughout the study. The main results were: 1) the interlaboratory CV of the prothrombin ratios obtained for each sample with all customary thromboplastins (5 different brands) was 15%, but was reduced to levels of 5.8 to 8.9 when using constant thromboplastin brands and batches; 2) the International Sensitivity Index (ISI) values obtained in the different labs were only slightly influenced by the use of different coagulometers; 3) comparable ISI values were obtained through direct calibration with the international reference material and through intermediate calibration with a locally selected standard; 4) use of INR values instead of percentage activity greatly reduced interlaboratory variability and significantly improved uniformity of anticoagulation level measurements, thus reducing the possibility of erroneous prescriptions. The Bologna exercise is therefore of educational value for laboratory and community doctors of the area in understanding and accepting the INR system.