Prevalence and Predictors of Subclinical Atrial Fibrillation in Hypertensive Elderly Patients Using Ambulatory ECG Monitoring

Abstract

Background: Atrial fibrillation (AF) is the most prevalent sustained cardiac arrhythmia worldwide and a major independent risk factor for ischemic stroke, heart failure, and systemic embolism. Hypertensive elderly patients represent a high-risk population in whom subclinical AF frequently goes unrecognized. Extended home-based ambulatory ECG monitoring improves detection but faces practical barriers in resource-limited outpatient settings. Single-session ambulatory ECG monitoring performed entirely within the outpatient department (OPD) offers a feasible, supervised, and immediately actionable alternative.
Objectives: To evaluate the diagnostic yield of a single OPD-based ambulatory ECG monitoring session (4–8 hours) for detection of subclinical AF in hypertensive patients aged 60 years and older, and to identify independent clinical and electrocardiographic predictors of AF detection in this setting.
Methods: This prospective cohort study enrolled 386 hypertensive patients aged 60–85 years presenting to the outpatient cardiology clinic. Each participant underwent a single continuous ambulatory ECG monitoring session of 4–8 hours duration within the OPD, using an AI-enabled single-lead adhesive patch electrode. Baseline clinical, echocardiographic, and laboratory data were collected on the same visit. AF episodes lasting >30 seconds were defined as clinically significant. Multivariable logistic regression identified independent predictors of subclinical AF detection.
Results: Subclinical AF was detected in 61 of 386 participants (15.8%) during the single OPD monitoring session. Independent predictors included frequent premature atrial contractions (PACs >200/session; OR 5.61, 95% CI 2.94–10.72), prolonged P-wave duration >120 ms (OR 4.28, 95% CI 2.27–8.08), left atrial dilatation ≥40 mm (OR 3.94, 95% CI 2.21–7.02), age ≥75 years (OR 2.71, 95% CI 1.52–4.83), obstructive sleep apnea (OR 2.31, 95% CI 1.22–4.37), and uncontrolled systolic hypertension >150 mmHg (OR 1.84, 95% CI 1.04–3.26). The AI detection algorithm achieved sensitivity 93.4%, specificity 96.6%, and AUC 0.974 versus expert electrophysiologist adjudication. Among AF-positive patients, 88.5% had CHA₂DS₂-VASc ≥2, and only 13.1% were receiving anticoagulation.
Conclusions: A single OPD-based ambulatory ECG monitoring session detects subclinical AF in approximately 1 in 6 hypertensive elderly patients and is both feasible and highly accurate with AI-assisted analysis. This approach represents a practical, scalable, and resource-efficient screening strategy for AF in outpatient cardiology practice.