Diagnostic accuracy of the PMcardio smartphone application for artificial intelligence-based interpretation of electrocardiograms in primary care (AMSTELHEART-1)

Jelle C L Himmelreich; Ralf E Harskamp

doi:10.1016/j.cvdhj.2023.03.002

Diagnostic accuracy of the PMcardio smartphone application for artificial intelligence-based interpretation of electrocardiograms in primary care (AMSTELHEART-1)

Cardiovasc Digit Health J. 2023 Apr 5;4(3):80-90. doi: 10.1016/j.cvdhj.2023.03.002. eCollection 2023 Jun.

Authors

Jelle C L Himmelreich^{1

2}, Ralf E Harskamp^{1

2}

Affiliations

¹ Department of General Practice, Amsterdam UMC location University of Amsterdam, Amsterdam, Netherlands.
² Personalized Medicine, Amsterdam Public Health, Amsterdam, Netherlands.

Abstract

Background: The use of 12-lead electrocardiogram (ECG) is common in routine primary care, however it can be difficult for less experienced ECG readers to adequately interpret the ECG.

Objective: To validate a smartphone application (PMcardio) as a stand-alone interpretation tool for 12-lead ECG in primary care.

Methods: We recruited consecutive patients who underwent 12-lead ECG as part of routinely indicated primary care in the Netherlands. All ECGs were assessed by the PMcardio app, which analyzes a photographed image of 12-lead ECG for automated interpretation, installed on an Android platform (Samsung Galaxy M31) and an iOS platform (iPhone SE2020). We validated the PMcardio app for detecting any major ECG abnormality (MEA, primary outcome), defined as atrial fibrillation/flutter (AF), markers of (past) myocardial ischemia, or clinically relevant impulse and/or conduction abnormalities; or AF (key secondary outcome) with a blinded expert panel as reference standard.

Results: We included 290 patients from 11 Dutch general practices with median age 67 (interquartile range 55-74) years; 48% were female. On reference ECG, 71 patients (25%) had MEA and 35 (12%) had AF. Sensitivity and specificity of PMcardio for MEA were 86% (95% CI: 76%-93%) and 92% (95% CI: 87%-95%), respectively. For AF, sensitivity and specificity were 97% (95% CI: 85%-100%) and 99% (95% CI: 97%-100%), respectively. Performance was comparable between Android and iOS platform (kappa = 0.95, 95% CI: 0.91-0.99 and kappa = 1.00, 95% CI: 1.00-1.00 for MEA and AF, respectively).

Conclusion: A smartphone app developed to interpret 12-lead ECGs was found to have good diagnostic accuracy in a primary care setting for major ECG abnormalities, and near-perfect properties for diagnosing AF.

Keywords: Atrial fibrillation; Cardiac arrhythmia; Digital health; Electrocardiogram; Primary care; Smartphone.