The utility of a baseline morning cortisol measurement, in the screening for adrenal insufficiency amongst non-critically ill hospital inpatients

Abstract

Background

Serum cortisol levels vary according to the daily circadian cycle, with peak levels seen in the morning. There is evidence in ambulatory patients that a morning cortisol level can predict an adequate cortisol response to Synacthen (tetracosactide) stimulation.

Aim

The aim of this study is to determine the utility of baseline analysis of the serum cortisol levels in the morning “baseline morning cortisol”, determined using a newer immunoassay, in the screening for adrenal insufficiency amongst non-critically ill hospital inpatients.

Methods

This is a retrospective cross-sectional cohort study.

Results

64 adult inpatients had undergone a short Synacthen test (SST) (measurement of serum cortisol levels 30 and 60 min after administration of tetracosactide) during the study period. 17 patients returned an abnormal SST result. The measured level of cortisol in the morning correlated to both the 30 min and 60 min stimulated cortisol values (r = 0.612, p < 0.001 and r = 0.639, p < 0.001). After inspecting the receiver operating characteristic curve, a cortisol concentration of 200nmmol/L measured in the morning was selected as a threshold for predicting the SST result. Using this cut off, sensitivity for predicting a normal SST was 100 %, specificity 56 %, positive predictive value 45 % and negative predictive value 100 %.

Conclusion

A morning cortisol measurement, determined using a newer immunoassay, is sufficient in most cases to screen for adrenal insufficiency amongst non-critically ill hospital inpatients. Use of clinical judgement in conjunction with single morning cortisol measurement is likely to reduce the need for SST testing amongst inpatients.