ALERT is the largest known multicenter, cluster-randomized trial to date evaluating an automated EHR-based notification system designed to address the undertreatment of significant valvular heart disease and accelerate time to treatment.

Tempus AI, Inc. (NASDAQ:TEM), a technology company leading the adoption of AI to advance precision medicine, today announced results from the ALERT (Addressing undertreatment and heaLth Equity in aortic stenosis and mitral regurgitation using an integrated ehR plaTform) trial, which were recently presented at the American College of Cardiology's 75th Annual Scientific Session & Expo. The study, conducted in collaboration with Medtronic, found that automated electronic clinician notifications (ECNs) integrated into the electronic health record (EHR) significantly improve the timely evaluation and treatment of patients with significant aortic stenosis (AS) and mitral regurgitation (MR).

Valvular heart disease is a leading cause of morbidity and mortality, yet it remains frequently undertreated. For patients with untreated symptomatic severe AS, mortality approaches 50% within just two years. Similarly, untreated severe MR carries a median survival of only five years. The ALERT trial was designed to determine if automated, AI-driven alerts could bridge this critical gap in care delivery. By leveraging the Tempus Next platform, which applies natural language processing to accurately extract findings from echocardiogram reports, the trial enabled real-time detection of significant disease and automatically delivered notifications with site-specific guideline-based care notifications directly to providers.

The ALERT trial included 765 clinicians and 2,016 echocardiograms across five U.S. health systems and 35 hospitals. The study met its primary endpoint, demonstrating that automated ECN alerts were superior to usual care in a win ratio analysis (win ratio 1.27; P = .007), meaning patients in the alert group were 27% more likely to be evaluated by the multidisciplinary heart team or receive a valve intervention than those in the usual care group. By delivering actionable data directly to providers, the system facilitated a 40% relative increase in life-saving valve procedures (13.4% vs. 9.6%) and a 27% increase in multidisciplinary heart team evaluations (22.7% vs. 17.9%) within just 90 days. These alerts effectively reduced clinical inertia, prompting earlier specialist referrals and ensuring patients received interventions within established benchmarks for timely care.

Beyond clinical efficiency, a central objective of the ALERT trial was to confront the persistent disparities that leave women, older adults, racial and ethnic minorities, and rural residents at higher risk of being undertreated. These findings suggest that EHR-integrated clinical decision support can serve as a powerful, scalable 'safety net,' standardizing care delivery to help ensure high-risk findings receive timely action regardless of a patient's demographics or care setting.