Letter Regarding “Comparison of 2 Assays for Measuring Serum Total Thyroxine Concentration in Dogs and Cats”

I am writing regarding the interesting article Comparison of 2 assays for measuring serum total thyroxine concentration in dogs and cats published in Volume 34, Issue 2 of JVIM in March 2020. The authors employ Passing-Bablok linear regression to measure the constant and proportional bias of the IDEXX Catalyst Total T4 Test (CTT4) when compared to a reference standard assay (Microgenics DRI TT4 assay; MTT4). This method requires a linear relationship between two data sets. Because the assumption of linearity was not met for the data from cats, piecewise regression analysis was performed, which established two linear relationships on either side of the breakpoint 1.80 μg/dL (Results, 3.2.1). Unfortunately, these two lines are not presented in the manuscript; instead, the invalid Passing-Bablok regression line is presented in Table 2 and Figure 3. As such, I have plotted the piecewise regression lines in Figure 1 of this letter. From this, it becomes evident that a CTT4 of 3.03 μg/dL may be equal to an MTT4 as low as 1.80 μg/dL. Assuming that the study is adequately powered and the CTT4 bias does not simply represent type I error, the IDEXX assay could, for example, result in a practitioner erroneously increasing the dose of methimazole in a hyperthyroid cat when, in fact, the dose was appropriate. Accordingly, I believe that Wolff et al.'s statement that For cats, the minimal bias present between the CTT4 versus MTT4 assays should not affect clinical decisions at concentrations < 10 μg/dL is misleading and requires correction. Furthermore, it should be stated in the conclusions that differences in CTT4 and MTT4 results may affect interpretation at lower, as well as higher, serum TT4 concentrations.

In addition, Wolff et al. state that, in dogs, At higher serum concentrations of TT4…a difference should be anticipated, with slightly higher concentrations expected with the CTT4 assay compared to the MTT4 assay. However, Figure 1 of their manuscript shows the opposite to be true (i.e., using the stated regression equation y = 0.103 + 0.860x, a CTT4 of 10 μg/dL is equivalent to an MTT4 of 11.51 μg/dL).

Yours sincerely,

Matthew K. Wun, Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Washington State University, Pullman, Washington, USA