Key result
Smartphone-based ECG AI matches POCUS for screening left ventricular dysfunction.
Why the study?
Point-of-care ultrasound requires specific equipment and skilled operators for heart function evaluation in emergency departments, motivating evaluation of an ECG-based mobile application for screening left ventricular dysfunction.
Does a mobile application analyzing ECG waveforms provide non-inferior diagnostic accuracy for screening left ventricular systolic dysfunction compared to point-of-care ultrasound in adult emergency department patients?
Observational
Does a mobile application analyzing ECG waveforms provide non-inferior diagnostic accuracy for screening left ventricular systolic dysfunction compared to point-of-care ultrasound in adult emergency department patients?
p-value: p==0.030
A smartphone application analyzing ECG waveforms can screen for left ventricular systolic dysfunction with accuracy comparable to point-of-care ultrasound.
May facilitate accessible LV dysfunction screening; leaves open prospective validation before clinical adoption.
PURPOSE: Point-of-care ultrasound (POCUS) is widely used for heart function evaluation in emergency departments (EDs), but requires specific equipment and skilled operators. This study evaluates the diagnostic accuracy of a mobile application for estimating left ventricular (LV) systolic dysfunction [left ventricular ejection fraction (LVEF) <40%] using electrocardiography (ECG) and tests its non-inferiority to POCUS. MATERIALS AND METHODS: Patients (aged ≥20 years) were included if they had both a POCUS-based EF evaluation and an ECG within 24 hours of their ED visit between January and May 2022, along with formal echocardiography within 2 weeks before or after the visit. A mobile app (ECG Buddy, EB) estimated LVEF (EF from EB) and the risk of LV dysfunction (LV-Dysfunction score) from ECG waveforms, which were compared to NT-proBNP levels and POCUS-evaluated LVEF (EF from POCUS). A non-inferiority margin was set at an area under the curve (AUC) difference of 0.05. RESULTS: =0.030, respectively). EF from EB was non-inferior to EF from POCUS, while LV-Dysfunction score narrowly missed the mark. A subgroup analysis of sinus-origin rhythm ECGs supported the non-inferiority of both EF from EB and LV-Dysfunction score to EF from POCUS. CONCLUSION: A smartphone application that analyzes ECG image can screen for LV dysfunction with a level of accuracy comparable to that of POCUS.
No takes yet. Share an insight, caveat, or question.
Kim et al. (2024) conducted an observational in Left ventricular systolic dysfunction. Mobile application (ECG Buddy) for estimating LVEF vs. Point-of-care ultrasound (POCUS) was evaluated on Diagnostic accuracy (AUC) for left ventricular systolic dysfunction (LVEF <40%) (p==0.030). A smartphone application estimating left ventricular ejection fraction from ECG waveforms was non-inferior to point-of-care ultrasound for screening left ventricular dysfunction (P=0.030).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: