Key Points
- The ALERT trial was designed to evaluate whether electronic clinician notification (ECN) alerts improve guideline-directed evaluation and treatment of significant AS and MR across multiple health systems.
- ECN improved the primary 90-day hierarchical composite of valve intervention or multidisciplinary heart team evaluation in ALERT (win ratio 1.27).
- Patients in the alert group were more likely to undergo valve intervention within 90 days, 13.4% vs 9.6%.
- Alerts also increased multidisciplinary team evaluation within 90 days, 22.7% vs 17.9%.
Severe aortic stenosis and mitral regurgitation are common, effective valve therapies are available, and yet treatment delays remain routine. ALERT tested whether an automated electronic clinician notification could move patients with significant valve disease to evaluation and treatment faster.
At the 2026 American College of Cardiology (ACC) Scientific Sessions, Dr. Wayne B. Batchelor presented the ALERT trial (NCT06099665), simultaneously published in The Journal of American College of Cardiology. ALERT was a multisystem, prospective, cluster-randomized clinical trial conducted across 5 US health systems and 35 hospitals. Clinicians ordering echocardiograms were randomized 1:1 to receive automated electronic clinician notifications for patients with significant aortic stenosis or mitral regurgitation, or to usual care without alerts. The primary endpoint was a hierarchical composite of time to surgical or transcatheter valve intervention, followed by time to multidisciplinary heart team clinic evaluation within 90 days.
The alert system was built to do one thing quickly: find patients with significant aortic stenosis or mitral regurgitation on echocardiography and push that information to the treating clinician through the electronic health record, along with guideline-based recommendations for next steps. The alerts were directed at clinicians, not patients.
In the primary modified intention-to-treat analysis, ALERT included 765 clinicians, 2,016 echocardiograms, and 1905 patients. The two groups were broadly similar at baseline, and the trial enrolled patients with both aortic stenosis and mitral regurgitation across inpatient and outpatient settings. By 90 days, the alert strategy outperformed usual care on the primary hierarchical composite of valve intervention or multidisciplinary heart team evaluation, with an overall stratified win ratio of 1.27 (95% CI: 1.05 – 1.54), p=0.007. The 90-day event rate for the composite endpoint was 24.3% in the alert arm versus 19.9% with usual care. Valve intervention occurred in 13.4% versus 9.6%, and multidisciplinary heart team evaluation in 22.7% versus 17.9%.
The treatment effect was similar across the two valve diseases studied. The win ratio was 1.29 in aortic stenosis and 1.23 in mitral regurgitation, with no evidence of heterogeneity by valve pathology or across prespecified subgroups such as age, sex, race, social deprivation, inpatient versus outpatient status, provider specialty, or rurality. The investigators noted that implementation varied across sites and required substantial institutional and technical support. They also emphasized that the alerts were clinician-directed rather than automated referrals, and that 90-day follow-up was too short to address longer-term clinical outcomes, particularly for mitral regurgitation.
In ALERT, an automated prompt inside the workflow was enough to move more patients from diagnosis to valve team evaluation and intervention within 90 days. As Dr. Batchelor put it, “A simple AI technology presents itself as a scalable digital health strategy to help us practice better and reduce the undertreatment of valvular heart disease, not to supplant docs, but just to give us an extra little nudge, so to speak, in the right direction.”