Abstract
Solid-phase first antibody (SP), liquid-phase double-antibody (DA), and preprecipitated double-antibody (PPT) separation methods have been compared in a radioimmunoassay (RIA) for cortisol in unextracted serum. Both double-antibody methods gave values for pools close to target values assigned by gas chromatography—mass spectrometry (GCMS) whereas the SP assay had a significant negative bias, p<0·001 (mean biases: SP —8%, DA —2%, PPT —3%). The SP and DA assays gave average values on patients' samples 12% and 4% lower than values by PPT. These differences could be attributed to different recoveries in the three methods. Between-assay precision of the DA and PPT systems was better than SP (mean CV: SP 11%, DA 7%, PPT 5%). This allowed sensitivity in the PPT assay comparable to that of SP and DA to be achieved, despite the less steep slope of the calibration curve. The DA and PPT assays have several practical advantages over SP. In addition, the PPT assay requires only a 1-hour incubation. It is concluded that the PPT separation system is the method of choice in terms of precision and practical convenience.