Clin. Lab. 12/2015 1947 Clin. Lab. 2015;61:1947-1952 ©Copyright ORIGINAL ARTICLE A Comparison of Three Widely Used Immunoassay Systems in Cortisol Measurement Nihal Yücel, Özlem Çakir Madenci, Aycan Bölük, Lale Köroğlu, Yusuf Temel, Asuman Orçun Dr. Lütfi Kirdar Kartal Training and Research Hospital, Clinical Biochemistry Department, Istanbul, Turkey SUMMARY Background: Interassay variability is one of the challenging issues of routine clinical laboratory practice. Com- mercial plasma cortisol immunoassays are also subject to thıs issue. In this study, we intended to evaluate the in- terchangeability of cortisol results of three widely used immunoassay systems. Methods: The cortisol values of 150 serum samples measured by three immunoassay systems, Beckman Coulter DXI 800, Roche Modular E170, and Siemens Immulite 2500, were compared. Results: A degree of proportional biases was observed between all three methods; DXI 800 showed the worst bias- es with the other two systems (slope values 0.67 and 0.77 with E170 and Immulite 2500, respectively). DXI 800 showed poor agreement with other methods (CCC: 0.83 and 0.87, respectively). There was a moderate agreement between E170 and Immulite 2500 (CCC: 0.92). Conclusions: All three methods showed a degree of variability among themselves. DXI 800 results were not inter- changeable with the other two systems. (Clin. Lab. 2015;61:1947-1952. DOI: 10.7754/Clin.Lab.2015.150539) Correspondence: Nihal Yücel Dr. Lütfi Kırdar Kartal Training and Research Hospital Clinical Biochemistry Department Orme Sitesi 18/7 334680 Çengelköy, Istanbul, Turkey Phone: +90 532 436 16 28 Email: yucel.nihal@yahoo.com _____________________________________________ Manuscript accepted June 10, 2015 KEY WORDS cortisol, immunoassays, transferability INTRODUCTION Serum cortisol, a major glucocorticoid hormone secret- ed by the adrenal cortex, is one of the most required tests in the routine laboratory. Basal values of serum cortisol or its response to stimulation or suppression tests are measured for a variety of clinical conditions in endocrinology and metabolism practice [1]. There are more than 30 immunoassay methods used in clinical laboratories and the discrepancy between meth- ods is still a matter of importance in routine practice. The transferability of the results is a central issue espe- cially when the laboratory immunoassay systems are switched to another. There are still numerous studies evaluating the comparability of the immunoassays [2,3]. The cortisol immunoassay method was first described 35 years ago [4] and adapted to automated systems. Se- rum cortisol is measured by immunoassays in routine laboratories. Although they are well defined and stan-