Problems with assays based on fibrin clot formation principle: analysis of multilaboratory data.

In order to investigate the inter-laboratory differences of coagulation assays, three different control plasmas were distributed for coagulation assays to 162 institutes. Three types of coagulation tests, that is, prothrombin time (PT), activated partial thromboplastin time (APTT), and fibrinogen (Fbg) assay were carried out by routine methods in each institute. In our laboratory, the Fbg concentrations were measured in two different plasmas using one instrument and nine reagents, and the Fbg concentrations were measured in ten control plasmas using three instruments and one reagent. Results were obtained as follows: (1) The coefficient of variance (CV) for control plasma with the highest value was significant in the PT assay and APTT assay. (2) Inter and intra-institute differences in PT and APTT assays were observed when the reagent or the instrument used was changed. (3) Inter and intra-institute differences in PT and APTT assays were also observed when both the reagent and the instrument were changed. (4) Inter and intra-institute differences in the PT and APTT assays were observed between the manual and the mechanical methods. (5) The CVs for the Fbg assays were the largest in the plasma at the lowest concentration of the three plasmas. (6) When control plasmas of three different Fbg concentrations were assayed, the number of the institutes which obtained a value greater than 3 standard deviations (SDs) was highest for the control plasma with the lowest concentrations of Fbg, and some of the instruments were not able to measure such a low Fbg concentration. (7) When the fibrinogen concentration was measured in various kinds of commercially-available control plasmas with a calibrated instrument, the values determined were different from the assigned ones and varied significantly from instrument to instrument. From the results, we concluded that standardization of the quality of control plasma, and its value assignment to the plasma should be established as soon as possible, and that a reliable standard fibrinogen assay should be developed.