A multicenter evaluation of immunoassays for follicle-stimulating hormone, luteinizing hormone and testosterone: concordance, imprecision and reference values.

BACKGROUND

Numerous laboratories in Italy use radioimmunoassay to determine concentrations of sex hormones (FSH, LH, testosterone). A comparison of assay methods is thus an important starting point for the achievement of universally accepted reference values.

AIM

To carry out an external quality assessment for FSH, LH, and testosterone.

MATERIALS AND METHODS

Fifteen aliquots from 5 serum pools were assayed in multiple replicates by 16 Italian laboratories with 5 automated immunoassays (Abbott Architect, DiaSorin Liaison, Perkin-Elmer AutoDelfia, Roche Elecsys, Siemens Immulite 2000), and 1 radioimmunoassay (Adaltis).

RESULTS

The variance was below 12% for FSH, between 11.61% and 14.76% for LH, and between 9.57% and 12.48% for testosterone. Assay precision was good, except for Elecsys at low concentrations of FSH and for Immulite at low concentrations of LH and testosterone. ARCHITECT showed a negative bias for FSH and LH and a positive bias for testosterone; Liaison a positive bias for LH; Elecsys a positive bias for FSH and a negative bias for testosterone; Immulite a positive bias for FSH; AutoDelfia a negative bias for FSH and a positive bias for testosterone. Reference ranges at the low end varied widely, even among laboratories using the same assay.

CONCLUSIONS

The analytical performances of widely used immunoassays for FSH, LH, and testosterone show a fair to strong degree of consistency. A careful evaluation of reference ranges by clinical and laboratory experts needs to be carried out, in order to reach a consensus.