Estimated incidence of previously undetected atrial fibrillation on a 14-day continuous electrocardiographic monitor and associated risk of stroke: comment

This Letter to the Editor refers to the article ‘Estimated incidence of previously undetected atrial fibrillation on a 14-day continuous electrocardiographic monitor and associated risk of stroke’ by William F. McIntyre et al. https://doi.org/10.1093/europace/euab324. We read with interest the article by McIntyre et al.1 who investigated the incidence of previously undetected atrial fibrillation (AF) on a 14-day continuous electrocardiogram (ECG) monitor and the associated risk of stroke. For these patients, who were aged >65 years with hypertension and a pacemaker, the findings demonstrated that a 14-day ECG detected AF >6 min in 3.1% and highlighted an associated increase in stroke risk [2.18%/year, adjusted hazard ratio (HR) 3.02]. As the findings have important clinical implications concerning the risk stratification of stroke, we propose that the authors also investigate other significant factors. The association between AF burden and increased stroke risk is corroborated by other recent clinical trials. Chu et al.2 conducted a retrospective cohort study of 152 patients, investigating pacemaker-detected AF burden, and associated ischaemic stroke or systemic thromboembolism risk (SSE). Using multivariate Cox regression, the authors determined that AF burden >6 min was a predictor of significantly higher SSE risk (propensity-adjusted HR, 6.75; P = 0.023) and reported a notably higher HR of 9.33 (P = 0.033) for any AF episode (including ≤6 min). Consequently, it would be interesting to see whether McIntyre et al.’s data on AF episodes ≤6 min was associated with increased incidence of SSE, allowing us to further evaluate the designated 6-min cut-off. In the stratification of stroke risk, CHA2DS2-VASc scoring is a clinical tool used to determine the need for oral anticoagulation in AF patients based on key risk factors. Nonetheless, in a study of 21 768 non-anticoagulated patients with pacemakers, Kaplan et al.3 reported that patients with a CHA2DS2-VASc score ≥5, despite not having AF, also crossed the actionable stroke risk threshold of >1%/year. While McIntyre et al. have provided a hazard ratio that is adjusted for the CHA2DS2-VASc score, we would recommend a separate analysis of the relationship between the CHA2DS2-VASc score and SSE in the patients without AF burden. This could provide highly valuable insights into the utility of CHA2DS2-VASc, and subsequent oral anticoagulation, in pacemaker populations without AF. Furthermore, the type of AF and ectopic beat burden may have bearing on the relationship with stroke risk, as observed by Cho et al.4 in their observational study of 10 000 AF patients. Atrial pre-mature beats were present in higher proportions in non-paroxysmal AF as opposed to paroxysmal (P = 0.001) in 24-h ECG monitoring, with a significantly increased risk of stroke in non-paroxysmal AF (adjusted HR 2.08, P = 0.001). Henceforth, we suggest that McIntyre et al. could explore SSE risk by AF type and ectopic beat burden—as these findings could deepen our understanding of the insidiousness of AF characteristics with respect to stroke, assisting prophylaxis, and risk management.