Heart · Diagnosis and imaging
AI stethoscopes in GP surgeries: what a 205-practice NHS trial found
The TRICORDER trial gave AI-enabled stethoscopes to NHS GP practices to help spot heart failure, atrial fibrillation and valve disease. Its main result was neutral: practices with the device didn't diagnose more heart failure overall. Many practices used it little. Where it was used, more diagnoses were recorded, but that comparison can't show the device caused them.
How far away is this?
Some of the device's AI software is FDA-cleared in the US, but this page is about its use in UK general practice, where the trial's main result was neutral. So we rate it on the trial evidence: In clinical trials.
Where is it approved?
| Region | Regulator | Status | Notes |
|---|---|---|---|
| United States | FDA | Cleared | Eko's Low Ejection Fraction Tool was cleared through a 510(k) (K233409, 28 March 2024), and its Murmur Analysis Software also has 510(k) clearance (K213794). We haven't confirmed these are the exact versions used in the UK trial.[5] |
| European Union | CE mark / EMA | Unknown | |
| United Kingdom | MHRA | Unknown | Used in NHS practices within the trial. We haven't checked the MHRA register. |
| Australia | TGA | Unknown |
Some of Eko's AI software is FDA-cleared in the US. The paper says the algorithms have regulatory approval, but we haven't verified the UK, EU or Australian records, so those are Unknown. "Unknown" means we haven't verified it, not that it's unavailable. Approval in one country isn't approval in another.
What is it?
A digital stethoscope records a 15-second heart trace (ECG) and heart sounds. Three algorithms then flag possible weak heart pumping, irregular rhythm or valve disease, so the GP can arrange proper tests.
What's the evidence?
205 UK practices were randomised (96 intervention, 109 control; over 1.5 million registered patients). Intention-to-treat analysis found no difference in heart failure detection (IRR 0.94, 95% CI 0.86 to 1.02).[1]
Evidence stage: In clinical trials · #c1
Intervention practices recorded 12,725 AI-stethoscope examinations. Examination with the device was associated with higher detection of heart failure, atrial fibrillation and valve disease. This is an observational comparison within the trial: practices and patients chosen for the device may differ from others, so it can't show the device caused the extra diagnoses.[7]
Evidence stage: In clinical trials · #c2
Researchers said use fell over time in some practices because the device added steps and wasn't well integrated with health record systems.[3]
Evidence stage: In clinical trials · #c3
The trial measured diagnoses recorded, not whether patients lived longer or stayed out of hospital. Extra flags can also mean extra tests for people who turn out to be fine. It was published online in The Lancet on 28 January 2026 and funded by the NIHR, the British Heart Foundation and Imperial Health Charity; the trial protocol says the manufacturer, Eko Health, was not involved in its design, conduct, analysis or reporting.[4]
Evidence stage: In clinical trials · #c4
What does it mean for me?
An AI stethoscope could come up at a GP appointment in some places. The lesson from this trial is that a good tool only helps if it's actually used. If you have breathlessness, swollen ankles or palpitations, tell your doctor, whatever tools they have. If you have sudden chest pain or severe breathlessness, call your emergency number.
What could I ask my doctor?
- Do you use any AI tools when listening to my heart?
- Would an ECG or heart scan be useful for my symptoms?
- If an AI tool flags something, what happens next?
These are general starting points, not a recommendation to choose any particular test or treatment.
If you're worried about your health right now, contact your doctor or local health advice line. In an emergency, contact your local emergency number.
Change history
- : First published.
- : Updated after independent fact-check and clinical review
- : Review label corrected: this page was not clinically reviewed. Earlier labels saying 'independent fact-check' and 'clinical review' were wrong; checks so far were AI-assisted source checks only.
General information only, not medical advice. Talk to your doctor about your situation. AI-assisted source check. Not clinically reviewed.
References
- The Lancet: TRICORDER trial. thelancet.com. Peer-reviewed randomised trial. https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(25)02156-7/fulltext
- ClinicalTrials.gov: NCT05987670. clinicaltrials.gov. Trial registry. https://clinicaltrials.gov/study/NCT05987670
- Imperial College London news (29 January 2026). imperial.ac.uk (2026). University news. https://www.imperial.ac.uk/news/articles/medicine/nhli/2026/ai-enabled-stethoscopes-show-promise-for-improving-diagnosis-of-cardiovascular-conditions-uk-trial-finds/
- BMJ Open: TRICORDER protocol (2025). bmjopen.bmj.com (2025). Peer-reviewed protocol. https://bmjopen.bmj.com/content/15/5/e098030
- FDA 510(k) K233409: Eko Low Ejection Fraction Tool. accessdata.fda.gov. Regulator. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K233409
- FDA 510(k) K213794: Eko Murmur Analysis Software (PDF). accessdata.fda.gov. Regulator. https://www.accessdata.fda.gov/cdrh_docs/pdf21/K213794.pdf
- Source for claim c2. doi.org. Source. https://doi.org/10.1016/s0140-6736(25)02156-7
Publisher shown as the website hosting each source. Dates are given where the source label includes them.