Assay of cortisol with a radioimmunoassay method calibrated by isotope dilution-mass spectrometry. A Nordic collaborative study.

A reference method for serum cortisol, based on isotope dilution-mass spectrometry (ID-MS), was compared with a modified commercial RIA method. The modification solely concerned the calibration standards used in the RIA method. These were replaced by a series of human serum samples, in which the concentration of cortisol had been determined by the reference ID-MS method. The samples were selected so as to cover the whole range of the standard curve. Serum samples from healthy, untreated subjects with cortisol concentrations 270-1134 nmol/l were analysed with the ID-MS calibrated RIA method in four laboratories, one in each of the four Nordic countries. The results from two of the four laboratories did not differ significantly from the ID-MS values (P less than 0.05), whereas the results from the other two laboratories differed slightly. The mean values based on results from all four laboratories were almost identical with the values obtained with the reference method. In addition, serum samples from 11 patients with endocrine disorders with cortisol concentrations 31-916 nmol/l were analysed in three of the four laboratories. In three of the samples significant differences were observed between the values obtained with the ID-MS and the ID-MS calibrated RIA method. The value obtained with the ID-MS calibrated RIA was however always more accurate than the corresponding value obtained with RIA with the use of a commercial calibration standard. Advantages of the present model in increasing the accuracy of routine analysis by RIA are discussed. The possibility is further discussed that the ID-MS calibrated RIA method may be used as a 'secondary reference method'.