Variability of cortisol assays can confound the diagnosis of adrenal insufficiency in the critically ill population

To compare the total plasma cortisol values obtained from three widely used immunoassays and a high pressure liquid chromatography (HPLC) technique on samples obtained from patients with sepsis. Observational interventional in the general intensive care unit of a metropolitan hospital Patients admitted to the intensive care unit with a diagnosis of sepsis and fulfilling criteria of systemic inflammatory response syndrome. Standard short synacthen test performed with 250 μg cosyntropin. Two of the three immunoassays returned results significantly higher than those obtained by HPLC: Immulite by 95% (95%CI 31–188%) and TDx by 79% (21–165%). The limits of agreement for all three immunoassays with HPLC ranged from −62% to 770%. In addition, by classifying the patients into responders and non-responders to ACTH by standard criteria there was concordance in all assays in only 44% of patients. Immunoassay estimation of total plasma cortisol in septic patients shows wide assay related variation that may have significant impact in the diagnosis of relative adrenal insufficiency.