Detection of paroxysmal atrial fibrillation : with special emphasis on intermittent ECG recording

Abstract

Atrial fibrillation (AF) is the most common sustained form of arrhythmia, affecting approximately 3 % of the adult population. AF may be symptomatic or asymptomatic. Asymptomatic paroxysmal atrial fibrillation (PAF) can be troublesome to diagnose, as most devices for arrhythmia detection provide only limited insight into heart rhythm. Regardless of the presentation of AF, it imposes the same risk of thromboembolic events.

The main purpose of these studies was to explore feasible methods for PAF detection. In Study I we evaluated the sensitivity and specificity of a handheld electrocardiographic (ECG) device by comparing concomitant 12-lead ECG registrations in 100 patients. Secondly, patients undergoing direct current (DC) cardioversion were asked to perform intermittent ECG recordings to investigate the ability to detect AF recurrence. Thirdly, 606 persons of the general public performed an ECG recording using the handheld device for feasibility assessment. Sensitivity and specificity of AF detection were 92% and 96 % respectively. Asymptomatic recurrence of AF after DC conversion occurred in 3/12 patients and was detected prior to standard follow-up. During screening, we detected previously undiagnosed AF in 1 % of the subjects.

In Study II, 22 patients with previously diagnosed PAF performed intermittent handheld ECG recordings, twice daily, for 30 days. Concomitantly they also performed a 24-hour continuous ECG registration. Using intermittent ECG recordings, 18 patients (82 %) were diagnosed with PAF episodes compared with 7 (32 %) using a continuous recording (p= 0.001).

In Study III, 249 patients without known AF who had recently suffered a cryptogenic ischaemic stroke/transient ischemic attack (TIA) were recruited. The patients underwent intermittent handheld ECG recordings twice daily for 30 days and performed a 24-hours continuous ECG recording in order to determine what modality detected the most patients with undiagnosed PAF. The diagnosis of PAF was set in 17 patients (6.8 %) all aged ≥ 65. Intermittent ECG recordings diagnosed 15 patients, continuous recordings 2 patients and only 3 patients were diagnosed by both methods.

In Study IV 174 patients, without known AF but with a recent ischemic stroke/TIA, underwent an echocardiographic examination using Tissue Doppler Imaging (TDI) to evaluate structural and functional parameters of the left ventricle and left atrium. PAF was diagnosed in 15 (8.6 %) patients. Patients with PAF had larger Left Atrial Volume Index (LAVI) (37.2±6.7 vs. 31.6±8.6 ml/m2) and lower intrinsic velocities (A′) in atrial (5.7±2.4 vs. 8.8±3.8) and ventricular (5.9±2.2 vs. 7.2±1.6) septa and higher LAVI/A ́ indices in ventricular (7.9±4.6 vs. 4.6±2.0) and atrial (8.8±3.8 vs. 5.7±2.4) septa. Reciever operating characteristic curve (ROC) analysis to detect later occurrence of AF was performed. The area under the curve (AUC) for LAVI was 0.71 and 0.78 for LAVI/A ́ in atrial septum.

Conclusions

Intermittent handheld ECG recordings accurately diagnose episodes of AF. Using intermittent ECG recordings over an extended period of time we can detect more episodes of AF compared with shorter continuous recordings in patients with and without known AF. Thus, the method appears to be an alternative to other modalities for AF screening. Echocardiographic investigations of LAVI, left atrial systolic contraction velocities (A′) and indices for these values may indicate which patients have underlying silent AF and might be used for identifying patients suitable for AF screening.