RESUMO
We performed a nationwide Inter-assay including 112 laboratories for the standardization of prothrombin time (PT) and thrombotest (TT). The data were expressed as seconds, percentile and INR. INR was expressed by 2 methods; Method I (conventional method): INR was expressed using each ISI assigned for reagent or reagent-instrument at the respective laboratories and Method II (local standardization method): INR was expressed using each reference curve created with INR assigned standard plasmas at the respective laboratories. (1) Sample distribution of PT as well as TT was the smallest with the data expressed by Method II followed by Method I and then by percentile. The data expressed by seconds was widely distributed and not useful for the standardization of PT and TT. (2) Even the sample distribution obtained by Method II was dependent on the different ISI of the reagents, as it was found that the larger the ISI of the reagents, the wider the distribution of data. (3) The difference between PT and TT of each test plasma was analysed by t-test. It was found that the difference was insignificant when both data were expressed by Method II, but significant when expressed by Method I, suggesting that PT and TT were interchangeable with the use of Method II. (4) Sample distribution of percentile expression and INR with the use of method II was compared. It was revealed that the sample distribution of INR was smaller than that of percentile. It was concluded that INR expressed by the local standardization method was most useful for the standardization of PT and TT.