Abstract:

Objective To evaluate the applicability of the Clinical and Laboratory Standards Institute(CLSI) guideline EP09-A3 in 2 chemiluminescence immunoassay systems determining serum total triiodothyronine(TT₃). Methods Totally，111 serum samples were determined for TT₃ and analyzed by System 1 and System 2. Analytical performance and method comparison were performed according to EP09-A3. Outliers were computed by extreme studentized deviate(ESD)，and 5 classified regression techniques were performed to evaluate the 2 methods and interpret biases. Results No outliers were determined in this study by visual judgement or ESD. The biases with 95% confidence interval(CI) of medical decision points(0.75 and 1.91 ng·mL^-1) were within the allowable total error(TEa)(±12.5%) by applying general linear regression，weighted least square regression，Deming regression and Passing-Bablock regression，and the outcomes were interpreted as type A or B which were acceptable. While the outcome of weighted Deming regression was interpreted as type C which may be judged as unacceptable. Conclusions The results of 2 systems determining TT₃ are comparable，and the analysis methods and interpretation of EP09-A3 are more scientific and reasonable.

Key words: EP09-A3, Total triiodothyronine, Method comparison, Medical decision point, Bias