SUN-410 Latent Class Analysis of Adrenal Insufficiency Symptoms Identifies Variation in Diagnostic Success and Outcomes - Suggesting Opportunities to Improve Patient Safety

Abstract Disclosure: A. Matti: None. G.L. Kinney: None. E.A. Regan: None. Adrenal insufficiency (AI) frequently has long delays in diagnosis. Diagnostic error has been defined as the failure to establish an accurate and timely explanation of the patient’s health problems. Patient safety initiatives emphasize the importance of accurate and timely diagnosis to prevent adverse outcomes. People experiencing adrenal insufficiency can experience unanticipated adrenal crisis and death before a diagnosis is established. Interpretation of patient-reported symptoms is key for diagnosis but not all patients experience disease symptoms in the same way. Physicians may utilize either truncated or classic patterns of symptoms to produce differential diagnoses and/or eliminate diagnoses considered rare, leading to diagnostic error and delay. Classic patterns of symptoms (for AI: fatigue, weakness, skin darkening for example) may not be sufficient to diagnose the condition without a comprehensive understanding the disease presentation. Certain patterns of symptoms or demographics may associate with greater success in diagnosis. We hypothesized that patient symptoms pre-diagnosis would impact the diagnostic outcome and studied a large cohort of people with adrenal insufficiency who had reported their experiences with diagnosis of AI. Methods: The MyAI registry was used to identify people (942) with a physician-diagnosis of AI. Participants were asked about their experiences with diagnosis of AI including what symptoms they associated with their ultimate diagnosis, their perception of factors that delayed diagnosis and whether they were diagnosed in adrenal crisis. Using latent class analysis to assess patterns of symptoms associated with their diagnosis, we identified six groups and labeled them based on group characteristics. Results: People reported difficulty with diagnosis in 551/ 843 (65%). Reasons cited: “Didn’t feel listened to” 27%, “Doctor thought diagnosis was too rare” 32%, “Patient not clear about symptoms” 2%, “Saw the wrong kind of doctor” 6%. “Was told that symptoms were not physical“ 46%.Diagnosis in crisis was reported in 52% overall.Latent class groups: Group 1 Classic symptoms n= 259, (82% female), 63% crisis diagnosis; Group 2 No symptoms, n= 34 (80% female), 26% crisis diagnosis, Group 3 Extended symptoms, n= 150 (91% female) 49% crisis diagnosis; Group 4 High symptom burden n= 100 (96% female) 52% crisis diagnosis; Group 5 Intensive extended symptoms (93% female) 68% crisis diagnosis; Group 6 Minimal classic symptoms n= 158 (80% female) 49% crisis diagnosis. Discussion:Men with fewer but classic symptoms have a better diagnostic experience. Diagnosis in crisis would represent diagnostic failure. Promoting a comprehensive list of symptoms associated with AI in physician education programs may improve outcomes and reduce adrenal crises. Presentation: Sunday, July 13, 2025