Background: Short Synacthen tests (SST) are both inconvenient and expensive, especially since the cost of tetracosactide recently increased 15-fold to over £45/ampule. A retrospective review was performed to see whether the number of SSTs could be reduced in our institution.
Methods: All 76 adult inpatient (mean age 67) and 106 adult outpatient SSTs (mean age 50) from a 12-month period were reviewed for indication and whether a preceding 0900 h cortisol was performed. The 30 min SST response and clinical interpretation were evaluated accordingly.
Results: Of all SSTs, only 12% of tests had a 30 min cortisol beneath 500 nmol/l, though 41% of these were interpreted as borderline. Only 32% of inpatient and 52% of outpatient SSTs had a preceding 0900 h cortisol measurement; the median of these was as high as 297 nmol/l, meaning a normal SST could be predicted with high certainty in many of these patients. 9% of 0900 h cortisol measurements were <70 nmol/l, thereby making an impaired SST a near certainty. 78% of inpatient SSTs were performed while investigating hyponatraemia, hypotension, falls or apparent hypoglycaemia, and all but one of these SSTs were interpreted as normal. 46% of outpatient SSTs were performed to exclude Addisons in the elective setting, and all of these were negative. 34% of outpatient SSTs were requested to complete endocrine workup in absence of any noted signs or symptoms of hypoadrenalism.
Discussion: Significant cost savings could be made by always requiring a 0900 h cortisol before performing SSTs. The 0900 h cortisol result should be interpreted in view of low pre-test probability of hypoadrenalism in most patients currently selected for SSTs. In addition, an SST is rarely needed in cases where a very low 0900 h cortisol is diagnostic. Strict protocols are required to ensure compliance to facilitate practice improvement.