Introduction: In 2017 the Roche cortisol assay underwent a generation update. The increased specificity of this assay was expected to lower cortisol results by approximately 30%. Based on a local consensus, the 30 min short Synacthen Test (SST) cortisol was lowered from 550 to 420 nmol/l. With the previous assay, a morning cortisol of ≥375 nmol/L gave a 99% PPV of an SST pass.
Aims: To compare the numbers of SSTs and the SST pass rates before and after the introduction of the new assay.
To establish the morning cortisol threshold likely to predict an SST pass.
Methods: 4 years of SST data was analysed from 2015 to 2019, encompassing 2 years before and 2 years after the new assay
Results: Total numbers of SSTs remained stable (634 vs. 663/2 years). A 30 min cortisol of 420 nmol/l resulted in similar pass rates with the new assay to 30 min cortisol of 550 nmol/l with the old assay (71%). A peak response of 380 nmol/l with the new assay gave similar pass rates to a peak of 500 nmol/l with the old assay (76%) and a PPV of 98.5%. To achieve a 99% PPV of passing the SST (based on a peak of ≥420 nmol/l with the new assay) morning cortisol was 390 nmol/l.
Conclusion: The more specific cortisol assay has resulted in uncertainty about thresholds for recommending and assessing SSTs. Kline et al. (2017) suggested a cut-off of 350 nmol/l for the peak SST response (reflecting the 30% lower cortisol results with the more specific assay). Most centres are using higher thresholds. We need to consider the potential of missing a diagnosis of adrenal insufficiency, best use of resources, and the adverse consequences of long-term steroid therapy for patients with normal pituitaryadrenal axes.