Triple cardiovascular disease detection with an artificial intelligence-enabled stethoscope (TRICORDER) in the UK: a cluster-randomised controlled implementation trial.

Artificial intelligence (AI)-enabled stethoscopes offer robust performance characteristics in point-of-care detection of heart failure, atrial fibrillation, and valvular heart disease (VHD).

人工智能（AI）赋能的听诊器在心脏衰竭、心房颤动和瓣膜性心脏病（VHD）的床旁检测中表现出强大的性能特点。

We conducted a pragmatic, cluster-randomised controlled implementation trial to determine the real-world effect and implementation challenges of AI-stethoscopes.

我们开展了一项实用的、群组随机对照实施试验，以确定人工智能听诊器在真实世界中的效果和实施挑战。

Methods

UK primary care practices were cluster randomised 1:1 to intervention (training and implementation in use of AI-stethoscopes in routine care) or control (routine care).

英国初级保健实践按1:1的比例被随机分为干预组（接受人工智能听诊器的培训和在常规护理中的实施）或对照组（常规护理）。

Given the nature of the intervention, masking of participants (practices, clinicians, and patients) was not feasible.

鉴于干预措施的性质，对参与者（实践、临床医生和患者）进行掩蔽是不可能的。

During cardiac examinations, the AI stethoscope recorded 15 s of single-lead electrocardiogram and phonocardiogram signals for input to three AI algorithms that returned binary predictions for the presence or absence of reduced left ventricular ejection fraction (≤40%), atrial fibrillation, and VHD (all with regulatory approval).

在心脏检查期间，AI听诊器记录了15秒的单导联心电图和心音图信号，输入到三个AI算法中，这些算法返回了二元预测结果，用于判断左心室射血分数降低（≤40%）、房颤和瓣膜性心脏病的存在或缺失（所有算法均获得监管批准）。

The primary endpoint was incidenceof any newly coded diagnosis of heart failure (all subtypes), expressed per 1000 patient-years (incidencerate ratio [IRR]), derived from a UK National Health Service Secure Data Environment.

主要终点是任何新编码的心力衰竭（所有亚型）的发生率，以每1000患者年为单位（发病率比[IRR]），数据来源于英国国家医疗服务体系安全数据环境。

A coprimary endpoint stratified detection of heart failure by place of diagnosis (community-based vs via hospital admission).

次要终点是根据诊断地点（社区与通过住院）对心力衰竭进行分层检测。

Secondary endpoints included atrial fibrillation and VHD detection rates, performance characteristics of the AI-stethoscope, use rates, and clinician-reported implementation barriers and enablers.

次要终点包括房颤和VHD的检出率、AI听诊器的性能特征、使用率，以及临床医生报告的实施障碍和促进因素。

Results

Between Oct 30, 2023, and May 22, 2024, 205 practices were randomly assigned (96 to the intervention arm [701 933 registered patients] and 109 to the control arm [851 242 registered patients]).

在2023年10月30日至2024年5月22日期间，共有205个实践被随机分配（干预组96个[注册患者701,933人]，对照组109个[注册患者851,242人]）。

Intervention practices recorded 12 725 patient examinations with the AI-stethoscope, across 972 clinical users.

记录的干预实践显示，共有972名临床用户使用AI听诊器进行了12,725次患者检查。

Intention-to-treatanalysis found heart failure detection did not differ between groups (IRR 0·94 [95% CI 0·86-1·02]); with no difference in community-based or hospital-based diagnoses (p>0·05).

意向治疗分析发现，两组间心力衰竭的检出率没有差异（IRR 0.94 [95% 置信区间 0.86-1.02]）；社区和医院的诊断也没有差异（p>0.05）。

interpretation

Implementation of an AI stethoscope in routine primary care did not significantly increase detection of heart failure or increase community-based diagnosis after 12 months of implementation.

在常规初级保健中实施人工智能听诊器并未显著提高心力衰竭的检出率，或在实施12个月后增加社区诊断。

AI stethoscope use was independently associated with significantly higher detection rates of heart failure, as well as atrial fibrillation and VHD.

人工智能听诊器的使用与心力衰竭、房颤和瓣膜性心脏病的检出率显著提高独立相关。

This randomised controlled implementation trial establishes a pragmatic design with randomisation that generates real-world data essential for understanding and overcoming the barriers to implementation of innovation in health care.

这项随机对照实施试验建立了一种实用的设计，通过随机化生成了对理解并克服医疗创新实施障碍至关重要的现实世界数据。

funding

National Institute for Health and Care Research, British Heart Foundation, and Imperial Health Charity.

国家卫生与护理研究院、英国心脏基金会和帝国卫生慈善机构。