Atrial Fibrillation Burden: An Update

The Need for a CHA2DS2-VASc-AFBurden Score

Kathryn D. Tiver; Jing Quah; Anandaroop Lahiri; Anand N. Ganesan; Andrew D. McGavigan


Europace. 2021;23(5):665-673. 

In This Article

Abstract and Introduction


Atrial fibrillation (AF) is an established independent risk factor for stroke. Current guidelines regard AF as binary; either present or absent, with the decision for anti-coagulation driven by clinical variables alone. However, there are increasing data to support a biological gradient of AF burden and stroke risk, both in clinical and non-clinical AF phenotypes. As such, this raises the concept of combining AF burden assessment with a clinical risk score to refine and individualize the assessment of stroke risk in AF—the CHA2DS2VASc-AFBurden score. We review the published data supporting a biological gradient to try and construct a putative schema of risk attributable to AF burden.


Atrial fibrillation (AF) has been demonstrated to be an independent risk factor for stroke in various populations.[1] The traditional biological explanation is of atrial stasis and thrombus formation, with risk of embolization.[2] It is biologically plausible that a greater burden of AF causes longer duration of atrial stasis and, therefore, increased risk of thrombus formation and subsequently stroke.

The traditional paradigm, as reflected in guidelines, suggests AF, regardless of type and burden, is a risk factor for stroke. The guidelines regard AF as a binary risk factor; either present or absent, irrespective of type, and solely use patient-level stroke risk factors, such as the CHA2DS2-VASc score, to inform the decision for anticoagulation.[3–5] This paradigm is represented graphically in Figure 1. Emerging data, however, suggest that the traditional binary stroke risk paradigm may be an oversimplification.[6,7]

Figure 1.

The traditional paradigm of stroke risk in AF, as reflected in the guidelines.2,3 Stroke risk (Y axis) is determined by CHA2DS2-VASc score alone, independent of AF burden (X axis). AF, atrial fibrillation.

It is important to note that the studies that inform these guidelines were conducted in patients with clinical AF.[8,9] Clinical AF is detected either due to symptoms, or, in the absence of symptoms, is likely to have a high burden as it is detected by chance. However, with the advent of wearable devices and increased information about long-term cardiac rhythm from implantable cardiac devices, the detection of subclinical and short duration AF is increasing.[10,11] Data from studies of device-detected AF have demonstrated consistently lower stroke risk compared to expected based on CHA2DS2-VASc score, challenging the paradigm of stroke risk equivalence across differing burdens of AF.

This raises the intriguing notion of AF burden as a driver for stroke risk, and recently, a dose-response relationship, supporting causality, has been reported to be plausible.[12] We use the available data to construct a putative schema of risk attributed to AF burden (Figure 2). This schema refines the overall stroke risk analysis to include AF burden, in addition to patient-level risk factors, to inform the decision for anticoagulation. The elusive aim, in order to aid in clinical decision-making, is to construct a dose-response curve, with a specific threshold over which anticoagulation is beneficial, to reduce embolic risk. Unfortunately, however, stroke risk has been assessed in studies of heterogeneous populations, with varying inclusion criteria and baseline risk, and therefore is not directly comparable.

Figure 2.

Putative schema describing relationship of AF burden, patient-level risk factors, and stroke risk. AF, atrial fibrillation.