Morning Serum Cortisol May Be Less Accurate for Diagnosing Adrenal Insufficiency in Hospitalized Patients, Particularly Those on Opioids

Abstract Introduction: The positive predictive value of a low morning serum cortisol to diagnose adrenal insufficiency (AI) is reported be >90%, which the basis for guidelines recommending as first-line test central AI. A level <3 µg/dL considered diagnostic. However, studies that established this cut-off were conducted primarily in outpatients, not hospitalized patients whom diurnal variation may disrupted. Studies suggest opioids suppress levels acutely, disrupt rhythm, and cause 8–50% chronically-exposed patients. impact hospitalization, particularly setting opioid use, on accuracy diagnostics AI unknown. We hypothesized values patients, especially those prescribed opioids, do accurately insufficiency, determined by corticotropin stimulation (CST) peak <18 µg/dL. Methods: performed retrospective analysis CSTs Mass General Brigham health system from 5/2015 9/2020. Opioid-exposed adults who underwent CST included if they had (5–9 AM) result. Control matched age, gender race no prescriptions within 30 days testing. Patients excluded tested outpatient or ICU setting, history cirrhosis pituitary disease, an elevated ACTH, oral estrogen, received oral, IV intraarticular glucocorticoids days. Results: 124 opioid-exposed 322 control mean (±SD) age 60.8±14.4 63.8±15.3y, 55.6% 45.0% female, respectively. Twenty-two (17.7%) 33 (10.2%) diagnosed with (p=0.031). Nineteen (15.3%) 22 (6.8%) (p=0.005). 36.8% (CI 19.1–59.0%) 50.0% 30.7–69.3%) In levels, daily morphine milligram equivalent duration exposure did differ between confirmed normal (p=0.13 0.38, respectively). Conclusion: Among suspected AI, have higher prevalence than controls. Serum accurate including more prevalent Given risks associated misdiagnosis caution should exercised relying diagnosis