[Validity of immunochemical assays for blood estradiol. Study conducted in 1994]

Since 1989 many kits have been commercialized for measuring estradiol by non-isotopic immunoassays which largely contributed to the diffusion of this parameter. Rapidly, in quality control schemes, a wide dispersion of results has been observed. In this paper, the kits' characteristics (precision, accuracy and detection limit) have been studied and compared to a reference method (gas chromatography-mass spectrometry). In addition, estradiol determinations in patients with various pathophysiological situations have been performed to underline the difficulties of interpretation of this analyte when clinicians compare results given by the different kits. This work demonstrates that precision and accuracy varied both with the kit and the level of concentration of estradiol and gives better results for high concentrations of estradiol. In seven out of the 12 kits, the limits of detection are in good agreement with those announced by manufacturers. On the contrary, the determination on patients sera gives results with a high variability with no systematic error.