Open AccessAvtomatizirovannyy elektrokhemilyuminestsentnyy metod opredeleniya kortizola v slyune dlya diagnostiki endogennogo giperkortitsizma sredi patsientov s ozhireniem
Author(s) -
Zhanna Evgenyevna Belaya,
А. В. Ильин,
Galina Melnichenko,
L Ya Rozhinskaya,
Н В Драгунова,
Larisa Dzeranova,
Н А Огнева,
С А Бутрова,
Е А Трошина,
Г С Колесникова,
И И Дедов
Publication year - 2011
Publication title -
ožirenie i metabolizm
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.154
H-Index - 5
eISSN - 2306-5524
pISSN - 2071-8713
DOI - 10.14341/2071-8713-4954
Subject(s) - medicine , confidence interval , intraclass correlation , receiver operating characteristic , likelihood ratios in diagnostic testing , reproducibility , odds ratio , immunoassay , gastroenterology , volunteer , coefficient of variation , nuclear medicine , antibody , immunology , chemistry , biology , chromatography , clinical psychology , agronomy , psychometrics
Nighttime salivary cortisol (NSC) has been suggested to be a useful diagnostic test for Cushings syndrome (CS). However, the reference range and cut-off value are assay-specific and discordant. The goal of this study was to assess the analytical performance of automated elecrochemiluminiscence immunoassay method (ECLIA) in CS. Ninety eight healthy volunteers and 123 obese patients including 45 proved to be CS provided salivary samples collected by them at 23:00 using Sallivette. Two hundred and five subjects collected salivary samples for two consecutive days and samples from 197 subjects were frozen to perform Enzyme-linked immunosorbent assay (ELISA). Obese patients underwent the 1-mg overnight dexamethasone suppression test (1-DST). CS was confirmed by the histologic diagnosis after surgical treatment or autopsy. The reference range for healthy volunteer has been set 0,5-9,4 nmol/l. Reproducibility was assessed in all subjects by a day-to-day variability and reflected by an intraclass correlation coefficient of 0,785. The cut-off value of 9,4 nmol/l has been suggested to differentiate CS among obese patients to achieve sensitivity of 84,4% (95%confidence interval 71,2-92,2%); specificity of 92,3% (95%CI 84,2-96,4%) and diagnostic odds ratio 65,1 (95% CI 20,4-207,6). Likelihood ratio positive was 11,0 (95% CI 5,0-23,9), with a likelihood ratio negative of 0,17 (95%CI 0,08-0,33). The comparison of the total areas under the ROC-curve for the measurement of NSC once, twice with mean level by ECLIA, the same samples by ELISA and 1-DST have not shown any statistically significant difference among the tests performance. Conclusion: Based on its remarkable reproducibility, easy noninvasive nature, automated assay and at least similar diagnostic performance, NSC measured by ECLIA on Cobas e601 is a preferable first-line screening test for CS.