Screening for paroxysmal atrial fibrillation in primary care using Holter monitoring and intermittent, ambulatory single-lead electrocardiography

Background: Timely detection of atrial fibrillation (AF) is important because of its increased risk of thrombo-embolic events. Single time point screening interventions fall short in detection of paroxysmal AF, which requires prolonged electrocardiographic monitoring, usually using a Holter. However, traditional 24-48 h Holter monitoring is less appropriate for screening purposes because of its low diagnostic yield. Intermittent, ambulatory screening using a single-lead electrocardiogram (1 L-ECG) device can offer a more efficient alternative. Methods: Primary care patients of ≥65 years participated in an opportunistic screening study for AF. We invited patients with a negative 12 L-ECG to wear a Holter monitor for two weeks and to use a MyDiagnostick 1 L-ECG device thrice daily. We report the yield of paroxysmal AF found by Holter monitoring and calculate the diagnostic accuracy of the 1 L-ECG device's built-in AF detection algorithm with the Holter monitor as reference standard. Results: We included 270 patients, of whom four had AF in a median of 8.0 days of Holter monitoring, a diagnostic yield of 1.5% (95%-CI: 0.4–3.8%). In 205 patients we performed simultaneous 1 L-ECG screening. For diagnosing AF based on the 1 L-ECG device's AF detection algorithm, sensitivity was 66.7% (95%-CI: 9.4–99.2%), specificity 68.8% (95%-CI: 61.9–75.1%), positive predictive value 3.1% (95%-CI: 1.4–6.8%) and negative predictive value 99.3% (95%-CI: 96.6–99.9%). Conclusion: We found a low diagnostic yield of paroxysmal AF using Holter monitoring in elderly primary care patients with a negative 12 L-ECG. The diagnostic accuracy of an intermittently, ambulatory used MyDiagnostick 1 L-ECG device as interpreted by its built-in AF detection algorithm is limited. •The diagnostic yield of Holter monitoring in screening for paroxysmal AF is low.•The diagnostic accuracy of ambulatory MyDiagnostick 1 L-ECG recordings is limited.•Cardiologists cannot reliably diagnose AF on ambulatory MyDiagnostick 1 L-ECG's.•Algorithm-positive ambulatory MyDiagnostick 1 L-ECG's are often uninterpretable.