Role of sampling times and serum cortisol cut-off concentrations on the routine assessment of adrenal function using the standard cosyntropin test in an academic hospital from Spain: a retrospective chart review

OBJECTIVES: Aiming to validate the use of a single poststimulus sampling protocol for cosyntropin test short standard high-dose test (SST) in our institution, our primary objectives were (1) to determine the concordance between 30 and 60 min serum cortisol (SC) measurements during SST; and (2) to evaluate the diagnostic agreement between both sampling times when using classic or assay-specific and sex-specific SC cut-off values. The secondary objectives included (1) estimating the specificity and positive predictive value of 30 and 60 min sampling times while considering the suspected origin of adrenal insufficiency (AI); and (2) obtaining assay-specific cut-off values for SC after SST in a group of subjects with normal hypothalamic–pituitary–adrenal (HPA) axis. DESIGN AND SETTING: This is a retrospective chart review study conducted at a Spanish academic hospital from 2011 to 2015. PARTICIPANTS AND INTERVENTIONS: Two groups were evaluated: (1) a main study group including 370 patients in whom SC was measured at 30 and 60 min during SST; and (2) a confirmative group that included 150 women presenting with a normal HPA axis in whom SST was conducted to rule out late-onset congenital adrenal hyperplasia. Diagnostic agreement between both sampling times was assessed by considering both classic (500 nmol/L) and assay-specific SC cut-off concentrations. RESULTS: Diagnostic agreement between both sampling times was greater when applying sex-specific and assay-specific cut-off values instead of the classic cut-off values. For suspected primary AI, 30 min SC determination was enough to establish a diagnosis in over 95% of cases, without missing any necessary treatment. When central AI is suspected, 60 min SC measurement was more specific, establishing a diagnosis in over 97% of cases. CONCLUSIONS: Sex-specific and assay-specific SC cut-off values improve the diagnostic accuracy of SST. For primary disease, a subnormal SC response at 30 min is a reliable marker of adrenal dysfunction. On the contrary, when central AI is suspected, 60 min SC measurement improves the diagnostic accuracy of the test.