[Determination of thyroxine in serum by a heterogeneous enzyme immunoassay: results of a joint trial].

This paper describes the evaluation of a heterologous enzyme immunoassay for the determination of total thyroxine in serum by a group of seven clinical chemical laboratories. The test follows the principles of the enzyme linked immunosorbent assay (ELISA) and uses peroxidase as a marker. The evaluation of analytical reliability yielded the following results within the analytical range from 39 unto 322 nmol/l: 1. Within-batch precision ranged from 3.1 unto 10.4% (coefficient of variation) with single analyses. 2. Between-batch precision ranged from 3.7 unto 20.4% with single analyses. 3. Between-laboratories precision ranged from 5.4 unto 6.8%. 4. Pure thyroxine, added to serum or thyroxine-free serum, gave recoveries between 93 and 120%. 5. Analysis of control sera gave results essentially comparable to the assigned values based upon radioimmunoassays. 6. Analysis of 288 clinical sera gave slightly higher results by the enzyme immunoassay than by the analogous radioimmunoassay from the same manufacturer. 7. Comparison with other methods of analysis (radioimmunoassays, competitive protein ligand assays, hormonal iodine assay) yielded partly comparable, partly higher results. 8. Comparison with the homogenous enzyme immunoassay (EMIT) led to comparable results. 9. Interference due to hyperlipemia or hemolysis was not observed. 10. There might be an interference in hyperbilirubinaemic sera, due to an as yet unknown factor. With respect to practicability the ELISA-test compares favourably with the analogous solid phase radioimmunoassay. The main differences are the absence of radioactive material and a longer shelf-live of reagents. Following the manual procedure the time taken to perform the enzyme immunoassay is slightly longer than for the analogous radioimmunoassay.