Triple Cardiovascular Disease Detection With an Artificial Intelligence-enabled Stethoscope (TRICORDER)

Heart failure (HF) is a condition in which the heart cannot pump blood adequately. It is increasingly common, consumes 4% of the UK National Health Service (NHS) budget and is deadlier than most cancers. Early diagnosis and treatment of HF improves quality of life and survival. Unacceptably, 80% of patients have their HF diagnosed only when very unwell, requiring an emergency hospital admission, with worse survival and higher treatment costs to the NHS. This is largely because General Practitioners (GPs) have no easy-to-use tools to check for suspected HF, with patients having to rely on a long and rarely completed diagnostic pathway involving blood tests and hospital assessment. The investigators have previously demonstrated that an artificial intelligence-enabled stethoscope (AI-stethoscope) can detect HF in 15 seconds with 92% accuracy (regardless of age, gender or ethnicity) - even before patients develop symptoms. While the GP uses the stethoscope, it records the heart sounds and electrical activity, and uses inbuilt artificial intelligence to detect HF. The goal of this clinical trial is to determine the clinical and cost-effectiveness of providing primary care teams with the AI-stethoscope for the detection of heart failure. The main questions it aims to answer are if provision of the AI-stethoscope: Increases overall detection of heart failure Reduces the proportion of patients being diagnosed with heart failure following an emergency hospital admission Reduces healthcare system costs 200 primary care practices across North West London and North Wales, UK, will be recruited to a cluster randomised controlled trial, meaning half of the primary care practices will be randomly assigned to have AI-stethoscopes for use in direct clinical care, and half will not. Researchers will compare clinical and cost outcomes between the groups.

Triple Cardiovascular Disease Detection with Artificial Intelligence-enabled Stethoscope (TRICORDER) is an open label, cluster randomised controlled trial. The aim is to determine whether use of an artificial intelligence-enabled stethoscope (AI-stethoscope) in UK Primary Care improves community-based detection of heart failure (HF), compared with usual care. 200 primary care practices in North West London and North Wales (UK) will be randomised to receive the AI-stethoscope (intervention arm) or continue with usual care (control arm). The intervention arm will use the AI-stethoscope in routine clinical practice. Outcomes will be measured using pooled primary and secondary care clinical and cost-data, as well as clinician questionnaires.

Heart Failure, Heart Valve Diseases, Atrial Fibrillation, Heart Murmurs, Congestive Heart Failure, Heart Failure With Reduced Ejection Fraction

cluster randomized controlled trial, artificial intelligence, digital health, cardiology, primary care

Receive 3-6 AI-stethoscopes (Eko DUO, Eko Health Inc, CA, USA) including artificial intelligence software for detection of: Reduced left ventricular ejection fraction <40% Atrial fibrillation Cardiac murmurs

Clinicians at practices in the intervention arm will be provided with one session of in-person training in use of the AI-stethoscope within 2 weeks of randomisation, including Delivery and setup Smartphone app installation and login Pairing of all clinician smartphones with all AI-stethoscopes in the same practice Demo of patient examination The AI-stethoscope will be used within its CE/UKCA-marked intended purpose. The clinical guidelines for use have been agreed by the NHS North West London Integrated Care System and Betsi Cadwaladr University Health Board Cardiovascular Executive Groups. Patients will be examined with the AI-stethoscope in accordance with these guidelines, and/or where stethoscope examination is deemed clinically appropriate. Patients will provide verbal consent for examination with the AI-stethoscope as per any physical examination performed by healthcare professionals for direct care, in accordance with UK law and General Medical Council guidelines.

Ratio of route to diagnosis of heart failure (co-primary) between emergency and community-based pathways

Difference in ratio of the incidence of coded diagnoses of HF via hospital admission-based versus community-based pathways.

Cost-consequence analysis (form of health economic evaluation) for diagnosis of atrial fibrillation, stratified by route to diagnosis. Presented in pounds sterling.

Cost-consequence analysis (form of health economic evaluation) for diagnosis of HFrEF, stratified by route to diagnosis. Presented in pounds sterling.

Cost-consequence analysis (form of health economic evaluation) for diagnosis of VHD, stratified by route to diagnosis. Presented in pounds sterling.

Health service utilisation for diagnostics e.g. rates of request for echocardiography, electrocardiography, primary care appointments. Collected from NHS organisation business intelligence repositories and UK Trusted Research Environments.

New implantation of cardiac resynchronisation therapy (CRT) and/or implantable cardioverter-defibrillator (ICD)

Patient-level sensitivity analyses will be performed for patients with abnormal Eko DUO predictions for HF, to identify direct associations between AI-stethoscope predictions and specific diagnostic codes for HF, AF and VHD

Inclusion Criteria: Primary care practices that care for adult patients and have the ability to request natriuretic peptide blood testing Primary care practices within the NIHR North West London Clinical Research Network or Betsi Cadwaladr University Health Board. Exclusion Criteria: Poor WiFi and/or mobile data connectivity within primary care consulting rooms No face-to-face patient consultations

Bachtiger P, Petri CF, Scott FE, Ri Park S, Kelshiker MA, Sahemey HK, Dumea B, Alquero R, Padam PS, Hatrick IR, Ali A, Ribeiro M, Cheung WS, Bual N, Rana B, Shun-Shin M, Kramer DB, Fragoyannis A, Keene D, Plymen CM, Peters NS. Point-of-care screening for heart failure with reduced ejection fraction using artificial intelligence during ECG-enabled stethoscope examination in London, UK: a prospective, observational, multicentre study. Lancet Digit Health. 2022 Feb;4(2):e117-e125. doi: 10.1016/S2589-7500(21)00256-9. Epub 2022 Jan 5.

Bachtiger P, Kelshiker MA, Petri CF, Gandhi M, Shah M, Kamalati T, Khan SA, Hooper G, Stephens J, Alrumayh A, Barton C, Kramer DB, Plymen CM, Peters NS. Survival and health economic outcomes in heart failure diagnosed at hospital admission versus community settings: a propensity-matched analysis. BMJ Health Care Inform. 2023 Mar;30(1):e100718. doi: 10.1136/bmjhci-2022-100718.