A fully automated direct assay of testosterone has been developed for
the Ciba Conning ACS:180 immunoassay system. We have evaluated this me
thod and compared specimen results with those from the Diagnostic Prod
ucts Corp, (DPC) and Medgenix direct assays and an in-house extraction
assay, Accuracy of the method was assessed by measuring GC-MS-targete
d serum pools, Within-assay imprecision was <6% and between-assay impr
ecision <9% over the concentration range examined, High concentrations
of lipid caused an increase in the measured testosterone but other po
tential interferents were without effect, Recovery was quantitative bu
t was affected by sex-hormone-binding globulin. The method was positiv
ely biased with respect to the DPC and Medgenix assays but negatively
biased with the extraction assay. Measurement of the GC-MS-targeted se
rum pools showed the assay to overestimate recovery by similar to 13%.
The ACS:180 assay is particularly attractive as a routine assay becau
se it is fully automated, obtains the first result in only 15 min, and
shows good assay performance.