Send to

Choose Destination
Stroke. 2014 Feb;45(2):520-6. doi: 10.1161/STROKEAHA.113.003433. Epub 2014 Jan 2.

Detection of atrial fibrillation after ischemic stroke or transient ischemic attack: a systematic review and meta-analysis.

Author information

From the Stroke and Vascular Research Centre, Institute of Cardiovascular Sciences (A.K., P.J.T., C.J.S.) and Centre for Biostatistics (A.V.), University of Manchester, Manchester Academic Health Science Centre, Salford Royal Foundation Trust, Salford, United Kingdom; Greater Manchester Comprehensive Stroke Centre, Department of Medical Neurosciences, Salford Royal Foundation Trust, Salford, United Kingdom (A.K., A.M., P.J.T., C.J.S.); and Institute of Cardiovascular and Medical Sciences, University of Glasgow, Western Infirmary, Glasgow, United Kingdom (J.D., K.R.L.).



Atrial fibrillation (AF) confers a high risk of recurrent stroke, although detection methods and definitions of paroxysmal AF during screening vary. We therefore undertook a systematic review and meta-analysis to determine the frequency of newly detected AF using noninvasive or invasive cardiac monitoring after ischemic stroke or transient ischemic attack.


Prospective observational studies or randomized controlled trials of patients with ischemic stroke, transient ischemic attack, or both, who underwent any cardiac monitoring for a minimum of 12 hours, were included after electronic searches of multiple databases. The primary outcome was detection of any new AF during the monitoring period. We prespecified subgroup analysis of selected (prescreened or cryptogenic) versus unselected patients and according to duration of monitoring.


A total of 32 studies were analyzed. The overall detection rate of any AF was 11.5% (95% confidence interval, 8.9%-14.3%), although the timing, duration, method of monitoring, and reporting of diagnostic criteria used for paroxysmal AF varied. Detection rates were higher in selected (13.4%; 95% confidence interval, 9.0%-18.4%) than in unselected patients (6.2%; 95% confidence interval, 4.4%-8.3%). There was substantial heterogeneity even within specified subgroups.


Detection of AF was highly variable, and the review was limited by small sample sizes and marked heterogeneity. Further studies are required to inform patient selection, optimal timing, methods, and duration of monitoring for detection of AF/paroxysmal AF.


atrial fibrillation; ischemic attack, transient; stroke

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Atypon
Loading ...
Support Center