Endocrine Abstracts

Background: Short Synacthen test (SST) is being increasingly used in various clinical conditions, mainly for completion purpose of ruling out adrenocortical insufficiency rather than actively suspecting it (for instance hypothyroidism with tiredness, type 1 diabetes with hypoglycaemia). The aim of our retrospective analysis was to assess if a random cortisol could predict the outcome of SST.

Method: Data were collected on all SST done at the endocrine unit over the last 18 months. Pre-test probability was scored as low-risk or high-risk based on indication for the test, analysed from clinical letters. Baseline cortisol done as part of SST was taken as random cortisol; 30 min post 250 μg i.v. Synacthen injection >550 nmol/l was considered as normal response.

Result: Of the 346 patients who had SST over the study period, 233 (67%) were identified as low risk. 221 of these (95%) had a normal SST response. A baseline cortisol of 400 nmol/l was found to predict adequate cortisol response to SST in low risk patients, with following specificity: >200–66; >250–91; >300–91; >350–91 and >400–100%. 100 of the 233 patients (43%) had a baseline cortisol of >400 nmol/l. Random cortisol of <140 nmol/l could predict inadequate SST response.

Among high risk patients (n=113), 58% had a normal SST response. Interestingly, random cortisol of 400 nmol/l predicted adequate response in SST with following specificity: >200–87; >250–91; >300–91; >350–94 and >400–100%.

Conclusion: A random (baseline) cortisol of 400 nmol/l can be considered as a marker of adequate adrenocortical function in patients with low pre-test probability. SST could have been deferred on 43% of the low risk patients. Clinical triaging of cases referred for SST, based on pre-test probability, and a random cortisol could help reduce the number of SST performed in metabolic units. This approach would help prioritize clinical appointments for high risk patients and streamline limited resources.