en.Wedoany.com Reported - Researchers from the Mayo Clinic and other institutions have developed an artificial intelligence (AI)-based screening model using 30 years of routine electronic health record (EHR) data, aimed at improving the identification of primary aldosteronism. Primary aldosteronism is a major cause of hypertension and often goes undiagnosed, increasing patients' risk of cardiovascular complications.
Primary aldosteronism occurs when the adrenal glands (small glands located atop the kidneys) secrete excessive amounts of the hormone aldosterone, leading to an imbalance of sodium and potassium levels in the body. Compared to patients with essential hypertension, these individuals face a higher risk of cardiovascular diseases, including stroke, coronary artery disease, atrial fibrillation, heart failure, and kidney disease.
Dr. Frank Lee, the lead researcher of the study from the Mayo Clinic in Rochester, Minnesota, noted that the true prevalence of primary aldosteronism is unknown but is estimated to affect up to 20% of hypertensive patients. Since effective treatments are available, early diagnosis can prevent future complications and reduce healthcare costs. The study will be presented on Saturday at the Endocrine Society's annual meeting, ENDO 2026, in Chicago, Illinois.
The Endocrine Society's 2025 publication, "Primary Aldosteronism: An Endocrine Society Clinical Practice Guideline," has already called for broader screening.
Researchers developed the AI screening model using the Mayo Clinic Platform, a privacy-preserving federated infrastructure with multimodal clinical data. The model analyzed de-identified data from over 22,000 patients collected between 1986 and 2025, including variables such as age, sex, ICD diagnoses related to hypertension and hypokalemia, systolic blood pressure measurements, serum potassium levels, and prescriptions for antihypertensive or potassium supplements. The model was then tested on data from 225,887 adult hypertensive patients. The study employed an XGBoost architecture, a type of machine learning library, to predict patients' risk of developing primary aldosteronism within 12 months before diagnosis.
Lee stated that the model demonstrates the feasibility of an AI-based screening approach. When researchers set a threshold to identify low-risk individuals, the model correctly flagged over 90% of primary aldosteronism cases, with a miss rate below 10%. Under this setting, approximately two-thirds of the study participants were identified as candidates requiring further screening.
Lee noted that when tested on hypertensive patients who had never been screened for primary aldosteronism, the model identified about two-thirds of patients needing further evaluation, pointing out that clinicians have struggled to effectively screen for the condition, and this tool offers a solution based on routine information from medical records.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
