Reliability of continuous single-lead patch ECG interpretation for atrial fibrillation detection: Insights from the NORSCREEN trial

Accurate interpretation of ECG recordings is essential for the diagnosis of atrial fibrillation (AF). Previous studies have demonstrated moderate reliability of single-point single-lead ECG in the setting of AF screening, which limits its utility for widespread use in early detection.

To evaluate the inter-observer agreement in the detection of AF by independent cardiologists interpreting continuous single-lead patch ECG recordings.

The Norwegian AF Self-Screening Trial (NORSCREEN) is an ongoing randomized controlled trial enrolling adults aged ≥65 years with at least one additional risk factor for stroke. Individuals with a prior AF diagnosis or anticoagulation therapy are excluded. Included participants are randomly assigned to long-term (7-day) ECG monitoring using the continuous single-lead patch ECG system ECG247 (Figure 1). All ECG recordings undergo analyses by an automated algorithm before review by a cardiologist (first ECG interpretation). Cases initially interpreted as AF by the cardiologist are further assessed by a Data and Safety Monitoring Board (DSMB) comprising two independent cardiologists (second ECG interpretation). In cases of disagreement, a third independent cardiologist adjudicates the final diagnosis. Inter-observer reliability was defined as the level of positive percent agreement between the first and second ECG interpretations. Minimum 30 s AF was required for an AF interpretation.

The 121 first interpretation of AF among the study participants were included in this analysis. The median age was 73 [interquartile range: 69–77] years and 48% of the study participants were female. The median duration of the ECG recordings were 7 days [interquartile range: 5–9 days]. The second ECG interpretation confirmed AF in 120 cases, corresponding to 99.2% agreement with the first interpretation (Figure 2).

The NORSCREEN trial demonstrate high inter-observer reliability in the interpretation of continuous single-lead patch ECG for AF detection. These results highlight the potential of continuous single-lead ECG monitoring to enhance AF diagnosis, thereby improving early detection and facilitating timely intervention in at-risk population.