Overview
- In the TRICORDER study across about 200 NHS GP practices and over 12,000 patients, detection within 12 months rose 2.33× for heart failure, 3.5× for atrial fibrillation and 1.9× for valve disease.
- The palm-sized sensor records heart sounds and a brief ECG, uploads encrypted data to the cloud for AI analysis, and returns a risk score to a practice smartphone in roughly 15 seconds.
- Findings were presented at the European Society of Cardiology congress in Madrid and highlighted by the British Heart Foundation, with the device reported as approved for use in the UK and its software classified as Class IIa.
- Real-world use lagged, with roughly 70% of participating practices scarcely using the device after a year due to time pressure, unclear workflows and uncertainty interpreting results.
- About two-thirds of AI-flagged suspected heart-failure cases were not confirmed on follow-up, prompting calls for confirmatory pathways and clinician guidance.