TOPLINE:

Automated electronic clinician notification (ECN) alerts improved the timeliness of guideline-directed evaluation and valve intervention among patients with clinically significant aortic stenosis (AS) or mitral regurgitation (MR) across multiple health systems in the US.

METHODOLOGY:

• Researchers conducted a cluster‑randomized superiority trial to examine whether fully automated electronic health record-integrated ECN alerts improved adherence to guideline‑directed care among patients with significant AS or MR.
• The intention-to-treat population included 765 clinicians who ordered 2016 echocardiograms in 1905 patients across five US health systems between August 2024 and September 2025. Clinicians were randomly assigned to receive ECN alerts identifying cases of significant AS or MR along with care recommendations (n = 376) or no alerts (usual care group; n = 389).
• The ECN alerts were generated through natural language processing that extracted structured and unstructured data from echocardiogram reports, with a previously validated accuracy of 98.6% for detecting severe AS.
• The primary endpoint was a hierarchical composite of surgical or transcatheter valve intervention followed by visit to a multidisciplinary heart team (MHT) clinic within 90 days of the index echocardiography.

TAKEAWAY:

• At 90 days, ECN alerts were superior to usual care for the primary hierarchical composite endpoint (overall stratified win ratio, 1.27; P = .007).
• Even after accounting for death as a competing risk, patients whose clinicians received ECNs had a higher probability of receiving care within 90 days than patients who received usual care (24% vs 20%; P = .027).
• In a separate analysis, ECN alerts were associated with a shorter time to valve intervention and evaluation at an MHT clinic (P = .005 for both).
• The benefits observed with ECNs were consistent across prespecified patient, provider, and care-setting subgroups.

IN PRACTICE:

“These findings suggest that EHR [electronic health record]-integrated clinical decision support may represent a scalable strategy to reduce the undertreatment of VHD [valvular heart disease] and improve timely access to specialized care,” the researchers of the study wrote.

SOURCE:

This study was led by Wayne B. Batchelor, MD, MHS, MBA, Inova Schar Heart and Vascular, Inova Health System in Fairfax, Virginia. It was published online on March 29 in JACC.

LIMITATIONS:

The chances of residual confounding related to site-level practice patterns, referral networks, and institutional resources could not be ruled out. Site‑level differences in alert wording, clinician awareness, and referral‑support infrastructure may have contributed to heterogeneity in the observed effects. Because the effect of ECN alerts depends on clinician engagement, factors such as alert fatigue, other clinical demands, and clinical inertia may have diminished their effectiveness. 

DISCLOSURES:

The study was funded by Tempus AI, Inc., and Medtronic plc. Several authors disclosed serving as speakers or consultants and receiving investigator-initiated grants, contracts, research and educational grants, and consulting or advisory fees from industry. Some authors disclosed being employees of and shareholders/stockholders in the funding agencies.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.