Diagnostic Accuracy Study of Morning Salivary Cortisone for Screening Adrenal Insufficiency in a Real-World Cohort.

CONTEXT: Morning serum cortisol is used to screen for adrenal insufficiency, but requires venepuncture, early morning clinic attendance, and may be affected by altered corticosteroid-binding globulin (CBG) levels. Salivary cortisone is a non-invasive, home-based alternative. OBJECTIVE: To assess the diagnostic accuracy of morning salivary cortisone for screening adrenal insufficiency and establish clinically relevant thresholds. DESIGN: Retrospective cross-sectional diagnostic accuracy study (2017 to 2024). SETTING: Single tertiary oncology and endocrinology center in the United Kingdom. PARTICIPANTS: Consecutive patients undergoing adrenocorticotropin hormone (ACTH) stimulation test for suspected adrenal insufficiency. INTERVENTION: Salivary sampling for cortisone and cortisol. MAIN OUTCOME MEASURES: Sensitivity and specificity of baseline salivary cortisone. RESULTS: A total of 517 ACTH stimulation tests in 449 participants were included; adrenal insufficiency was confirmed in 104 tests (20.1%). Baseline salivary cortisone accurately predicted adrenal insufficiency (area under the receiver operating characteristic curve [AUC] 0.93, 95% CI [0.91-0.95]), similar to serum cortisol (0.93, 0.90-0.95; p = 0.67) and superior to salivary cortisol (0.87, 0.82-0.91; p = 0.00063). A baseline salivary cortisone rule-out threshold ≥703 ng/dL (19.5 nmol/L) (sensitivity 0.97 [95% CI 0.92-0.99]; specificity 0.59 [0.54-0.63]) and a rule-in threshold <180 ng/dL (5.0 nmol/L) (sensitivity 0.58 [0.48-0.67]; specificity 0.97 [0.95-0.98]) provided the highest net benefit. An intermediate threshold <389 ng/dL (10.8 nmol/L) (sensitivity 0.83 [0.74-0.89]; specificity 0.87 [0.83-0.90]) identified patients for whom glucocorticoid cover should be considered while awaiting confirmatory testing. CONCLUSION: Baseline morning salivary cortisone performs at least as well as serum cortisol for adrenal insufficiency screening.