Background: Short Synacthen tests (SST) are inconvenient and expensive, especially since the cost of tetracosactide recently increased to over £45/ampule. Historic literature on the minimum 9 am cortisol required to avoid a SST ranges from 243 to 500 nmol/l and individual endocrinologist practice varies greatly.
Methods: In an audit of all 182 SSTs from 1 year at our institution, either a previous 0900 h cortisol was recorded or the basal cortisol was measured in the SST. The 30 min stimulated response was evaluated and the documented clinical interpretation noted.
Results: Of 79 patients with a previous 0900 h cortisol level, 71 measured above our laboratorys normal 0900 h lower limit of 119 nmol/l. 66 of these had a stimulated cortisol above 500 nmol/l and were interpreted as normal. The remaining five patients all had a stimulated cortisol above 400 nmol/l, but only one was judged to require commencement of hydrocortisone and two were already taking hydrocortisone. Of 103 patients without a previous 0900 h cortisol, 95 had a basal cortisol in the SST above 119 nmol/l. 91 of these had a stimulated cortisol above 500 nmol/l. The remaining four patients all had a borderline response above 400 nmol/l, but only one was judged clinically to require steroid replacement. In summary, of 166 patients with basal or 0900 h cortisol above 119 nmol/l, only two were judged to require commencement of hydrocortisone therefore the SST did not change management in 99% of cases.
Discussion: Defining normality in measured adrenal function is notoriously difficult. Borderline SSTs are rarely acted upon, and it is usually doubtful if these patients require steroid replacement. It therefore does not seem clinically relevant to detect these cases. We therefore argue that most SSTs could be avoided by measuring a 0900 h cortisol level if it is within the local laboratory true 0900 h normal range.