Evaluation of an automated enzyme-linked fluorescent assay for thyroxine measurement in cat and dog sera.

Measurement of total thyroxine (T4) is the first testing step in the work-up of thyroid disease in small animals. We evaluated an enzyme-linked fluorescent assay (ELFA) as an in-house method to measure T4 in cats and dogs. We compared the T4 concentration in sera of 122 cats and 176 dogs measured by the ELFA with an enzyme immunoassay (EIA) to assess the concordance of the 2 methods. Bias of the ELFA in cats was -11.4% and in dogs 1.4%. Using Bland-Altman plots, limits of agreement were -81.5 to 58.7% in cats and -71.4 to 74.4% in dogs. Imprecision was calculated for both methods. Intra- and interassay coefficients of variation (CVs) of the ELFA in feline sera were 0.7 and 3.4% and of the EIA 7.6 and 15.7%, respectively. Intra- and interassay CVs of both ELFA and EIA in canine sera were <9.5%. Reference intervals for the ELFA method were established and were 13.3-49.5 nmol/L for cats and 10.1-42.9 nmol/L for dogs. Accuracy of the EIA and ELFA was scored by assessing if the measured T4 value would identify the expected T4 range (low, normal, or elevated) of patients, based on history, clinical presentation, other diagnostic means, and response to therapy. This was possible for 75 cats and 50 dogs. Both methods yielded acceptable results, but the EIA was more accurate compared to the ELFA (percentage of true-positives in cats and dogs: EIA: 97% and 100%; ELFA: 92% and 94%).