AI-enabled stethoscope trial aims to transform heart disease detection in primary care

A UK-based study is testing whether an artificial intelligence (AI)-enabled stethoscope can improve early detection of cardiovascular disease in community settings.

The study, named ‘Triple cardiovascular disease detection with an artificial intelligence-enabled stethoscope (TRICORDER): design and rationale for a decentralised, real-world cluster-randomised controlled trial and implementation study’ is the world’s first large-scale, decentralised cluster-randomised controlled trial of its kind, as published in NIH.

The device is designed to detect three major cardiovascular conditions:

Left ventricular systolic dysfunction (a cause of heart failure)

Atrial fibrillation (irregular heart rhythm)

Cardiac murmurs (linked to valvular disease)

Up to 200 primary care practices across urban North West London and rural North Wales will take part. Practices are being randomised either to continue usual care or to use the AI-enabled stethoscopes, with clinicians free to use them at their own discretion, supported by regional clinical guidelines.

Unlike conventional trials, TRICORDER has no patient-level inclusion or exclusion criteria. Instead, outcomes will be tracked through pooled NHS primary and secondary care records, supplemented by clinician surveys.

The study has two primary endpoints–

Increase in coded incidence (detection) of heart failure.

Change in the ratio of hospital-based vs community-based diagnosis of heart failure.

Secondary outcomes include detection rates of atrial fibrillation and valvular heart disease, cost-effectiveness, and prescription of guideline-directed therapies.

The trial has received ethical approval from the UK Health Research Authority and is registered at ClinicalTrials. 

Researchers believe the device, if proven effective, could revolutionise heart disease detection in primary care, offering earlier, community-based diagnosis and reducing reliance on hospital admissions.