Performance evaluation of the Access FT3 and FT4 assays, comparison with Immulite and AxSym, and the relationship to...

Different FT3 and FT4 assays report significantly different results. We compared the distribution of FT3 and FT4 in a cohort of Swiss patients measured with DxI 800, AxSYM, and Immulite 2000.

TSH, FT3, and FT4 values were measured in 1,938 serum samples. Patients were classified on the basis of their TSH values as low, normal, and high. For each class of TSH values, concordances of FT3 and FT4 results were determined among the three assays.

For low TSH values in all three assays FT3 (FT4) concordance of DxI - AxSYM, DxI--Immulite, and AxSYM--Immulite was determined as 83.1%, 76.2% 68.5% (60.8%, 74.6%, 83.1%), for normal TSH as 89.2%, 79.0%, 75.3% (83.9%, 85.5%, 83.1%) and for elevated TSH as 78.0%, 86.0%, 78.0% (84.0%, 90.0%, 90.0%), respectively. Low FT4 concordance rates with DxI 800 were mainly caused by its FT4 upper reference limit of 14.1 pmol/L. Using a cut-off of 16.1 pmol/L concordances with AxSYM and Immulite were improved to 77.7% and 86.9% (low TSH), 92.5% and 96.2% (normal TSH), and 90.0% and 92.2% (high TSH). Low FT3 concordance rates with Immulite were caused by its low FT3 upper reference limit of 6.29 pmol/L as 11.6% of patient samples with normal TSH value showed unusually elevated FT3 results.

We showed an overall good concordance of FT3 and FT4 results, when stratified according to corresponding TSH values and the appropriate reference range is used. However, our data also show that problems of interpretation of results based on numerical values have yet not been solved.