Journal of Innovative Clinical Trials and Case Reports

Abstract

Background: Conventional acoustic auscultation is limited by subjectivity and marked inter-observer variability. Valvular heart disease is missed by standard auscultation in more than half of cases, and in the United Kingdom over 70% of heart failure diagnoses are made only after an unplanned hospital admission despite prior consultations at which symptoms were already present [1-3]. Digital stethoscopes coupled to artificial-intelligence (AI) classification algorithms have been proposed as a means of narrowing this diagnostic gap, but prospective clinical evidence in hospitalised internal medicine and cardiology populations remains limited.

Methods: We describe the clinical rationale, design, and protocol of the Phase 1 investigational clinical validation of an AI-assisted digital stethoscope (STETHO-HUS), conducted within the Service de Médecine Interne and Département de Cardiologie of the Hôpitaux Universitaires de Strasbourg (HUS). Consecutive adult inpatients and outpatients undergo standardised auscultation with the investigational device, benchmarked against transthoracic echocardiography and high-resolution CT/pulmonary function testing as reference standards, under an approved Comité de Protection des Personnes (CPP) protocol.

Findings: Enrolment is ongoing; no diagnostic accuracy data are yet available. An interim analysis is planned after 100 complete cases.

Interpretation: This report should be read as a description of the clinical rationale and study protocol for an investigational device. Diagnostic accuracy, clinical utility, and impact on patient management remain to be established through the ongoing, adequately powered validation study.