Technical Note: Significant Positive Bias in Lower Concentrations but Negative Bias in Higher Concentrations in...

Testosterone is the principal male sex hormone and is secreted primarily by the testes. In most clinical laboratories testosterone is routinely measured for diagnosis of male hypogonadism or androgen excess in females using FDA approved immunoassays. We compared testosterone values measured by Beckman access immunoassay with those measured by a reference LC-MS/MS method.

Testosterone was measured in 36 patients using left over serum or plasma specimens by both Beckman immunoassay on the DXI 800 analyzer and a reference LC-MS/MS method.

We observed overall significant negative bias of approximately 31.9 % when testosterone values obtained by the reference LC-MS/MS method were plotted in the x-axis and the corresponding testosterone values using the immunoassay in the y-axis, as regression equation was y=0.6887x+38.81 (n=36). The corresponding Deming regression was y=0.6639x+34.7163. However, in eight specimens with low testosterone concentrations, immunoassays significantly overestimated testosterone concentrations.

Immunoassays may underestimate the true testosterone concentration in males but overestimate in females with low testosterone concentration. Therefore, for diagnosis of hypogonadism in males and androgen excess in females, testosterone values obtained by Beckman Access immunoassay on the DXI 800 analyzer should be interpreted with caution.