[Commutability of reference materials for coagulation factor VIII and factor IX activity on three measurement...

To evaluate the comparability of measurement results for coagulation factor VIII (FVIII)and factor IX (FIX) activity and the commutability of reference materials on different measurement systems.

The study was performed according to CLSI guideline EP30 and China health standard WS/T 356-2011. Clinical samples with different levels of FVIII and FIX which covered over the clinical analytical range, five lots of homemade reference materials (F20140601-F20140605) and a coagulation reference material (SSCLOT4) provided by NIBSC were detected for FVIII and FIX activity on three popular measurement systems in China, which including Stago STA-R Evolution, IL ACL TOP700 and Sysmex CA7000 automatic coagulation analyzers using supplementary reagents. The results between measurement systems were analyzed pairwise. To evaluate the comparability, the linear regression and the biases between the results of clinical samples from two measurement systems were calculated. The comparability was evaluated by the regression coefficient and the biases inside the acceptable range. After eliminated outliers from the results, linear regressions were run again and the 95% confidence intervals were calculated. The commutability of the homemade reference materials and NIBSC reference material were evaluated by comparing the results with the limits of the intervals.

The ranges of FVIII and FIX level of clinical samples were 0.5%-218.0% and 1.6%-156.5%, which covered the sample levels in routine work and fit the requirements for commutability evaluation. The square of correlation coefficients (R²) of measurement results of clinical samples for FVIII and FIX activity assays were 0.89-0.94 and 0.81-0.93. The proportions of outliers were all less than 10%. The comparability of measurement results of FVIII and FIX in different measurement systems was acceptable.According to the acceptable criteria for bias, the measurement results of 42, 41 and 45 clinical samples for FVIII and 44, 42 and 41 clinical samples for FIX were used in the commutability evaluation for homemade reference materials and NIBSC reference material after deleting the outliers. The five lots homemade reference materials and NIBSC reference material were all within the 95% prediction intervals from the linear regression of clinical samples' results.

Good correlation and comparability are found for FVIII assay and those for FIX assay are acceptable. All homemade reference materials and NIBSC reference material are commutable for the three measurement systems.