Prospective blinded evaluation of smartphone-based ECG for differentiation of supraventricular tachycardia from inappropriate sinus tachycardia

Abstract

Introduction Supraventricular tachycardias (SVT) are often difficult to document due to their intermittent, short-lasting nature. Smartphone-based one-lead ECG monitors (sECG) were initially developed for the diagnosis of atrial fibrillation. No data have been published regarding their potential role in differentiating inappropiate sinus tachycardia (IST) from regular SVT. If cardiologists could distinguish IST from SVT in sECG, economic health care burden might be significantly reduced. Methods We prospectively recruited 75 consecutive patients with known SVT undergoing an EP study. In all patients, four ECG were recorded: a sECG during SVT and during sinus tachycardia and respective 12-lead ECG. Two experienced electrophysiologists were blinded to the diagnoses and separately evaluated all ECG. Results Three hundred individual ECG were recorded in 75 patients (47 female, age 50 ± 18 years, BMI 26 ± 5 kg/m2, 60 AVNRT, 15 AVRT). The electrophysiologists’ blinded interpretation of sECG recordings showed a sensitivity of 89% and a specificity of 91% for the detection of SVT (interobserver agreement κ = 0.76). In high-quality sECG recordings (68%), sensitivity rose to 95% with a specificity of 92% (interobserver agreement of κ = 0.91). Specificity increased to 96% when both electrophysiologists agreed on the diagnosis. Respective 12-lead ECG had a sensitivity of 100% and specificity of 98% for the detection of SVT. Conclusion A smartphone-based one-lead ECG monitor allows for differentiation of SVT from IST in about 90% of cases. These results should encourage cardiologists to integrate wearables into clinical practice, possibly reducing time to definitive diagnosis of an arrhythmia and unnecessary EP procedures. Graphical abstract A smartphone-based one lead ECG device (panel A) can be used reliably to differentiate supraventricular tachycardia (panel B) from inappropriate sinus tachycardia when compared to a simultaneously conducted gold-standard electrophysiology study (panels C, D).

Reliability of single-lead electrocardiogram interpretation to detect atrial fibrillation: insights from the SAFER Feasibility Study

“…In comparison, in the present study of single-lead ECGs cardiologists agreed on the diagnosis of 86.1% of ECGs (1,587 out of 1,843 ECGs). However, the proportion of normal ECGs included in this study was substantially lower than in the SAFE AF Screening Trial (less than 1% in this study, versus 𝜅 = 0.76 when using single-lead ECGs from the same patients(17).The previously reported levels of agreement for the diagnosis of AF from single-lead ECGs have varied greatly between studies: fair agreement was observed by Kearley et al(18) (𝜅 = 0.28); moderate agreement was observed byLowres et al…”

“…However, the proportion of normal ECGs included in this study was substantially lower than in the SAFE AF Screening Trial (less than 1% in this study, versus 93% in SAFE), so the simple level of agreement is not directly comparable. Similarly, in a study of the diagnosis of supraventricular tachycardia in hospital patients, an almost perfect agreement of 𝜅 = 0.97 was observed in interpretation of 12-lead ECGs, compared to a substantial agreement of 𝜅 = 0.76 when using single-lead ECGs from the same patients (17). The previously reported levels of agreement for the diagnosis of AF from single-lead ECGs have varied greatly between studies: fair agreement was observed by Kearley et al (18) (𝜅 = 0.28); moderate agreement was observed by Lowres et al (19) (weighted 𝜅 = 0.4); substantial agreements were observed by Poulsen et al (12) (𝜅 = 0.65) and Kearley et al (18) (𝜅 = 0.76); and almost perfect agreements were observed by Desteghe et al (11)(𝜅 = 0.69 to 0.86), Koshy et al (20) (𝜅 = 0.80 to 0.83), Wegner et al (13) (𝜅 = 0.90), and Racine et al (21)(𝜅 = 0.94).…”

Consumer-Led Screening for Atrial Fibrillation: Frontier Review of the AF-SCREEN International Collaboration

“…On the other hand, the presence of ventricular pacing or atrial flutter often leads to misclassification of AF as sinus rhythm. 32 These issues will limit accuracy of AF algorithms in consumer-facing devices.…”

Taquicardia sinusal: Caso clínico

“…En la actualidad el mundo ha evolucionado y la tecnología ha evolucionado por lo que realizó con un grupo de estudio de 75 pacientes de un área específica de salud con un teléfono moderno inteligentes que fue de utilidad diagnóstica que beneficiara a los médicos a dar un diagnóstico certero y seguro y confiable (3).…”