Echocardiographic risk stratification for atrial fibrillation detection using 7-day holter monitoring in patients without history of atrial fibrillation: a prospective study

Abstract

Diagnosis of subclinical atrial fibrillation (AF) is challenging due to its intermittent nature. While prolonged Holter monitoring improves detection, current guidelines lack clear criteria for selecting patients for extended rhythm monitoring, limiting resource allocation. This study aimed to identify echocardiographic parameters to develop a risk score guiding the optimal selection of patients for 7-day Holter monitoring. We prospectively enrolled patients aged ≥ 60 years without a history of AF between 2022-2025. Clinical and echocardiographic parameters were collected, followed by 7-day Holter monitoring every 6 months. The primary endpoint was incident AF, defined as an episode lasting > 30 seconds on Holter or documented on a 12-lead ECG. Multivariable Cox proportional hazards regression identified independent predictors to construct a simplified risk score. The model was internally validated via bootstrapping. Among 769 patients (mean age 71.7±5.5 years), incident AF was detected in 48 (6.2%). Multivariable analysis identified four independent echocardiographic predictors: LA reservoir strain < 35% (aHR 2.31, p=0.02), LAVI/A' ≥ 4 (aHR 2.80, p=0.001), TR Vmax > 2.5 m/s (aHR 2.41, p=0.004), and average PA-TDI > 105 ms (aHR 1.94, p=0.03). A simplified 4-point risk score was developed (AUC 0.73, p<0.001), with each 1-point increase associated with a 2.36-fold higher risk of AF (aHR 2.36, 95%CI 1.79–3.09; p<0.001). Diagnostic performance analysis showed that a score > 0 yielded a sensitivity of 89.6% and NPV of 98.1% for ruling out AF, while a cut-off ≥ 2 maximized specificity (75.7%) for identifying high-risk candidates. Based on these thresholds, the score effectively stratified patients into three tiers with a stepwise increase in AF incidence (p<0.001): low-risk (0–1 point), intermediate-risk (2 points), and high-risk (3–4 points), with cumulative AF incidences of 3.4%, 11.3%, and 22.6%, respectively. Results remained consistent in sensitivity analysis addressing lead-time bias. This novel echocardiographic risk score provides effective risk stratification for subclinical AF and establishes a robust framework for personalized triage. It optimizes resource allocation by prioritizing 7-day Holter monitoring for high-risk patients to maximize diagnostic yield. Conversely, shorter monitoring or standard care may be sufficient for low-risk individuals, while intermediate-risk patients warrant close clinical follow-up.Central illustration ECHOLTER-AF study