RESUMO
BACKGROUND: Measurement of estradiol (E(2)) plays a critical role in the diagnosis and clinical management of reproductive disorders. The challenge for all currently available direct methods for measuring E(2) is to provide accuracy and precision across a wide dynamic range. METHODS: We describe the development and multi-site performance evaluation of a direct E(2) assay on the Architect i2000. Assay performance and method comparisons were performed by testing specimens from men, healthy women with regular menstrual cycles, and post-menopausal women using the Architect assay and isotope dilution, gas chromatography-mass spectrometry (ID/GC-MS). Reference intervals were established by testing prospectively collected daily blood draws from 42 healthy women, 72 postmenopausal women and 101 males. RESULTS: No unexpected cross-reactivity or interference was observed for over 40 compounds tested. Recovery was 100+/-10% in the presence of estrone and estriol. Functional sensitivity (%CV<20%) was <15 pg/ml.(1) The imprecision of the assay was <7.1% (total CV), <2.5%, and <2.3% for control sera containing 45, 190, and 600 pg/ml estradiol, respectively. The assay had a correlation of y=1.033 x+0.3156, r(2)=0.99, n=131 compared to ID/GC-MS. Reference intervals for the current Architect Estradiol assay are reported. CONCLUSIONS: Format changes resulted in dramatic improvement in the performance and accuracy of this direct, fully automated assay. The assay is standardized by ID/GC-MS. The assay is clinically useful for serum concentrations from 15 to >4000 pg/ml.