SUN-430 Assessing Adherence to the Inpatient Cosyntropin Stimulation Test Protocol for Diagnosis of Adrenal Insufficiency: A Quality and Patient Safety Report

Abstract Disclosure: M.D. Lundholm: None. M. Ortega Abad: None. P.P. Rao: None. Introduction: Adrenal insufficiency (AI) is a life-threatening condition requiring timely identification and management. The 3-point cosyntropin stimulation test (CST) is used to assess for AI, yet is challenging to coordinate: a basal cortisol blood draw is followed by intravenous or intramuscular 250 mcg of cosyntropin (synthetic ACTH) with subsequent cortisol blood draws at 30 and 60 minutes. We aim to analyze our system’s adherence to the inpatient CST procedure to ensure the accurate and timely diagnosis of AI. Methods: This retrospective quality and patient safety project includes all adult patients who had an inpatient CST performed between September 2020 - February 2022 across the Northeastern Ohio region of our healthcare system. Charts were reviewed for cross-reacting steroid use (prednisone, methylprednisolone, hydrocortisone), albumin level, and adherence to the CST procedure. Results: We include 514 patients, mean age 60.4 ± 15.7 years, 53.5% (N=275) female, who collectively had 554 inpatient CSTs, of which 14.4% (N=80) were consistent with AI. Procedural issues were identified in 30.1% (N=167) of all CSTs: 20.0% (N=111) had blood draw collection or timing issues, 6.1% (N=34) ACTH was not given, or given after the test ended, 4.3% (N=24) were on cross-reacting steroids, and 2.3% (N=13) had lab errors. Additionally, 18.2% (N=101) tests were performed in the setting of albumin <2.5 g/dL, only 5.0% (N=5/101) checked free cortisol levels, yet 85.1% (N=86/101) reached cortisol ≥12.6 mg/dL to exclude AI. Procedural issues or low albumin interfered with 44.2% (N=245) of all tests. Of those results consistent with AI, 57.5% (N=46/80) were performed incorrectly or with low albumin without free cortisol, potentially over-diagnosing AI. Of those results ruling out AI, 2.5% (N=12/474) had cross-reacting steroid interference risking a missed diagnosis of AI. In all, 10.4% (N=58) of tests were unable to diagnose or rule out AI. Endocrinology was consulted in 30.5% (N=169) of CSTs with a trend toward higher consult use in cases with procedural issues or low albumin (50.3% vs 41.6%, p=0.057). Discussion: At a large academic institution, over 30% of inpatient CSTs were performed incorrectly, most often due to blood draw timing issues, omitted cosyntropin, and interfering steroid use. Low albumin remains a challenge and free cortisol is rarely used in part due to long turnaround time. Despite this, the majority of patients with low albumin were still able to exclude AI with total serum cortisol. Overall,10.4% of CST results were uninterpretable, which can delay a critical diagnosis of AI, or cause over-diagnosis of AI. Given the importance of prompt steroid intervention, but also the harms of unnecessary steroid use, we must prioritize performing the CST correctly the first time. Further study is planned to determine the impact of education and safety interventions to address these issues. Presentation: Sunday, July 13, 2025