Evaluation of the analytical performance of a point-of-care analyzer for the measurement of feline serum thyroxine concentration in comparison with a chemiluminescence enzyme immunoassay

Background: Total thyroxine (TT4) measurement is used to assess thyroid status in cats. Objectives: The aim of the prospective study was to evaluate the analytical performance of the point-of-care analyzer (POCA) Immuno AU10V using the v-T4 test kit for feline TT4 measurement. Additionally, method comparison with a benchtop analyzer (IMMULITE 2000) was done. Methods: Validation included linearity, inter- and intra-assay precision, precision near the lower limit of quantification (LloQ), and interference testing for hemoglobin, lipid, and bilirubin. Correlation and bias were assessed. Results: Linearity was given within the dynamic range. Coefficients of variation (CV) were <= 4% near the LloQ as well as for intra-and inter-assay precision. No interference was observed for lipid and bilirubin, while hemoglobin caused a negative bias of 28%. Method comparison included 74 samples within three TT4 concentration ranges (0.5-3.7, >3.7-5.13, >5.13-8 mu g/dL). Correlation between POCA and reference method was excellent (r(s) = 0.95) with a slight proportional bias of 4.5%. TEobs was between 7.0% and 9.8%. Despite substantial agreement, discordant results on thyroid status occurred in 15% of samples. Conclusions: The analytical performance of the POCA was excellent, as was its correlation with the reference method. Except for the interferent effect of hemoglobin, the TEobs was <TEa for all analyses. Analysis of severely hemolytic samples is not advised. However, the relatively small dynamic range of the POCA precludes quantitative analysis of samples with TT4 >8 mu g/dL, and de novo reference intervals need to be established.