Basal serum cortisol levels predict a normal response to the Synacthen stimulation test in hospitalised patients.

Abstract

The short Synacthen test (SST) is widely used to assess the hypothalamus-pituitary-adrenal axis in the outpatient setting. However, in the inpatient setting, technical difficulties to adhere to the protocol may pose a challenge for using this test. To find the most suitable basal serum cortisol (BSC) cut-off for predicting an adequate response to the SST in non-critically inpatients without conducting the actual test. Information was retrieved retrospectively from medical files of 197 patients who had had a 250 μg SST between the years 2000 and 2016 at the Shaare Zedek Medical Center. BSC, electrolytes, creatinine, thyroid-stimulating hormone, blood counts and blood pressure values were evaluated for a correlation with the results of the SST. A BSC cut-off of 280 nmol/L provides a negative predictive value of 94% for adrenal insufficiency (AI). Using a cut-off of 380 nmol/L increases the sensitivity to 96% and yields a negative predictive value of 95.8%. In this study, we found two suitable BSC cut-offs for predicting an adequate response to the SST in hospitalised patients. We suggest using the lower cut-off (280 nmol/L) for patients with a low level of suspicion for AI and using the higher cut-off (380 nmol/L) for patients with a higher level of suspicion. A BSC above this cut-off makes the diagnosis of AI very unlikely and precludes the need for a Synacthen test.