Background: Ischaemic stroke is a leading cause of morbidity and mortality worldwide with around 100,000 individuals suffering a stroke in the United Kingdom each year. Around a fifth of these events are cardioembolic in origin, often due to arrhythmias such as atrial fibrillation (AF) or atrial flutter (AFl). Although national guidelines currently recommend prolonged ECG monitoring to assess for a cardioembolic source, the optimal length of cardiac monitoring remains uncertain and there is no consensus as to the minimum duration of atrial fibrillation that constitutes a risk of thromboembolism. Studies have indicated that the insertion of an implantable cardiac monitor (ICM) can improve the detection of atrial arrhythmia. However, there is also evidence that extending monitoring beyond 7 days duration is unlikely to be cost-effective. We aimed to assess the current use of ambulatory ECG (AECG) in patients admitted with acute ischaemic stroke at a large teaching hospital.
Methods: A retrospective observational study was performed on all patients who were admitted to Leeds Teaching Hospitals NHS Trust with a diagnosis of acute ischaemic stroke between January 2014 and December 2015. The number of AECGs performed, the total recording time and the frequency of supraventricular tachyarrhythmia were recorded.
Results: A total of 1,706 patients were admitted with confirmed ischaemic stroke over the 2-year period. The mean age was 73.5 years and there was no significant difference in gender (p=0.483). In patients with known AF/AFl, 6.57% underwent an AECG following their stroke, accounting for a total of 33 ambulatory monitors and a mean recording time of 67.6 hours. In patients without known atrial arrhythmia, 0.02% were found to have AF/AFl on 12-lead ECG. Of the remaining patients, 72.1% underwent AECG (accounting for 1,332 monitors with a mean recording time of 63.0 hours) with 20.9% undergoing multiple AECGs. Paroxysmal AF or AFl was seen in 5.6% and 0.4% of all AECGs respectively. The detection of paroxysmal AF increased as the length of AECG recording increased and this finding is highlighted in Figure 1. Interestingly, persistent AF or AFl was seen in 7.3% and 0.4% of all AECGs respectively.
Conclusions: This study confirms that the detection of atrial arrhythmia improves as the length of cardiac rhythm monitoring increases. However, there remains significant variability in the length and number of AECGs requested. Additionally, a significant proportion of AECGs are performed in patients who do not require cardiac rhythm monitoring, either because they have a history of atrial arrhythmia or because they have persistent AF/AFl at presentation. A more detailed clinical assessment at presentation could help to reduce the demand of AECGs in this patient population. As a consequence of this study, we have now recommended that for patients who present with acute ischaemic stroke to our Trust, a 72-hour AECG is performed as standard if there is a suspicion of a cardioembolic source. More prolonged AECG, including the insertion of an ICM, is only considered if a patient has suffered more than one ischaemic stroke in separate cerebrovascular territories.