Israel Medical Association Journal

Background: Electrocardiogram (ECG) may detect atrial fibrillation (AF), but the true rate of ECG performance is unknown.

Objectives: To explore the performance rate of ECG testing in patient populations at high risk for AF in a real-world cohort and to explore the incidence of newly diagnosed AF.

Methods: This retrospective observational cohort study included de-identified data of members of Maccabi Health Services older than 65 years, excluding patients with prior AF and other cardiac diseases. Patients were followed between 1 January 2016 and 30 September 2020. The number of ECGs performed in an outpatient setting for each patient was reported.

Results: In total, 211,515 patients (59.8% female) were included. The mean age was 70.3 ± 6.6 years, with a mean CHA₂DS₂-VASc score of 2.6 ± 1.0. During the study period, over half of the patients (n=112,340; 53.1%) did not undergo any ECG tests, 51,644 patients (24.4%) had one ECG, 24,914 patients (11.8%) had two, while 22,617 patients (10.7%) had more than three. Of the patients referred for an ECG, 67,433 (81.1%) underwent ECG testing within 2 months following the referral. The median time from ECG referral to ECG testing was 5 days.

Conclusions: Most patients older than 65 years had no ECG tests within a 5-year period. However, when referred to an ECG test, most patients complied within a median of 5 days.

Background: Since the introduction of the electrocardiogram (ECG) in 1902, the fundamentals of ECG data acquisition, display, and interpretation in the clinical arena have not changed much.

Objectives: To present a new method to enhance and improve acquisition, analysis, and display of the standard ECG.

Methods: We performed ECG enhancement by superimposition and summation of multiple standard ECG cycles of each lead, by temporal alignment to peak R wave and voltage alignment to an improved baseline, at the T-P segment.

Results: We enhanced ECG recordings of 504 patients who underwent coronary angiograms for routine indications. Several new ECG features were noted on the enhanced recordings. Examination of a subgroup of 152 patients with a normal rest 12-lead ECG led to the discovery of a new observation, which may help to distinguish between patients with and without coronary artery disease (CAD): namely, a spontaneous cycle-to-cycle voltage spread (VS) at the S-T interval, normalized to VS at the T-P interval. The mean normalized VS was significantly greater in those with CAD (n=61, 40%) than without (n=91, 60%), 5.61 ± 3.79 vs. 4.01 ± 2.1 (P < 0.05).

Conclusions: Our novel method of multiple ECG-cycle superimposition enhances the ECG display and improves detection of subtle electrical abnormalities, thus facilitating the standard rest ECG diagnostic power. We describe, for the first time, voltage spread at the S-T interval, an observed phenomenon that can help detect CAD among individuals with normal rest 12-lead ECG.