Abstract P008: Repeatability of Ectopic Beats From 48-hour Ambulatory Electrocardiography: The Atherosclerosis Risk in Communities (ARIC) Study

Abstract

Background: Ectopic beats, defined by premature atrial contractions (PACs) and premature ventricular contractions (PVCs), are one of the most common findings on the electrocardiogram (ECG). Excess ectopic beats on ambulatory ECG monitoring are associated with atrial fibrillation (AF), stroke and death in symptomatic patients and in population-based studies. However, the repeatability of ECG-enumerated ectopic beats has not been sufficiently examined in a general population. Objectives: Characterize the repeatability of PACs and PVCs in a community-based population of older adults. Methods: Ambulatory, 48-hour ECG was measured in a subsample of the Atherosclerosis Risk in Communities Study (ARIC; 2011-2013) at two locales. Participants were randomly invited to participate in the repeatability study (n=92). Following an identical, standardized protocol, the participants wore a Holter monitor for two, 48-hour periods separated by a mean of 38 days (range: 11, 130 days) and completed a symptom questionnaire during each period. Each visit included new electrodes and Holter placement. All participants wore monitors for 48 hours and had recordings with <1% noise. Participants with AF (n=4) or paced rhythm (n=2) were excluded (total n=86; mean age 74.6 years; 56 females, 55 black). Visit-specific quartiles and 80th percentile cut-points for the total counts of PACs and PVCs over the 48 hour monitoring period were used. Between-visit agreement was quantified using the weighted kappa for quartiles and kappa for 80th percentile cut-points. Results: About 17% of the participants reported symptoms of dizziness, palpitations, or chest pain while wearing the monitor. During the 48-hour recordings, all participants had ≥1 PAC count and 2 participants had a 0 PVC count. The overall median (interquartile range (IQR)) was 266 counts (IQR: 80, 1,692.5) for PACs and 223 counts (IQR: 17.5, 1,085.5) for PVCs. The median between-visit difference (visit 2 - visit 1) was -12.5 counts for PACs (IQR: -154, 65) and 0 counts for PVCs (IQR: -78, 320). The between-visit, weighted kappa for count quartiles was 0.70 (95% confidence interval (CI): 0.61, 0.80) for PACs and 0.68 (95% CI: 0.58, 0.78) for PVCs. The kappa for counts >80th percentile was 0.56 (95% CI: 0.34, 0.78) for PACs and 0.71 (95% CI: 0.52, 0.90) for PVCs. Conclusion: Between-visit agreement is high to moderate for 48-hour, Holter-based PACs and PVCs quantiles, according to common interpretations for kappa coefficients. When paired with pending examination of how agreement varies over shorter duration recordings within the 48-hour monitoring period, the present results can facilitate epidemiologic and clinical applications in which knowledge of measurement error, variability and misclassification are needed to inform risk assessment.