Standardisation of thrombin generation test--which reference plasma for TGT? An international multicentre study

We have previously shown that standardisation and normalization of results improve the intercentre variability of the calibrated automated thrombin generation test (TGT). We suspected that the source of reference plasma (RP) might be a contributing factor to variability and compared 5 commercial RP and a RP provided by the NIBSC, in an international, multicentre study. The detailed composition of the 6 tested plasma samples was determined in the Haemostasis Laboratory in Lyon. The lot to lot consistency, intra-assay, inter-assay variability were calculated for all tested plasmas. The RP and 3 plasma samples (a normal control, a hypocoagulable and a hypercoagulable plasmas) were tested over 6 days, in 5 European centres. Results were normalised against each of the tested RP and intercentre variability of results was compared. All laboratories used the same reagents. Before normalization, the inter-centre variability was 19.8 to 27.3%. After normalization, we observed a significantly improved inter-laboratory variation with all tested RP, despite differences between them. These results clearly demonstrate that the inter-centre variability of TGT can be significantly reduced by using a reference plasma normalization, and that certain RP have a better capacity to reduce this variability than others.