See related article, p 1210. Post-hospital evaluation for the cause of ischemic stroke has gained increased attention in the last decade, particularly the search for paroxysmal atrial fibrillation. Initial studies of Holter monitoring documented rates of atrial fibrillation of ≈5% or 1 in 20 patients.1 Further, prolonged monitoring for as long as 1 week appeared to increase detection rates, with a range of 5.7% to 7.7%.1 When devices became available to increase monitoring ≤21 days, detection rates over 20% were documented.2 A 2014 systematic review of detection of atrial fibrillation after stroke found an overall rate of detection of 11.5%; however, detection rates varied based on patient selection, device selected, timing of device placement, duration of monitoring, and diagnostic criteria for paroxysmal atrial fibrillation.3 Two randomized trials of prolonged monitoring after ischemic stroke, both published after the 2014 systematic review, found higher rates of atrial fibrillation with longer duration of monitoring. The 30-Day Cardiac Event Monitor Belt for Recording Atrial Fibrillation after a Cerebral Ischemic Event (EMBRACE) trial found a 90-day rate of 16.1% in patients undergoing 30 consecutive days of monitoring versus 3.2% of patients undergoing 24 hours of monitoring.4 In the Cryptogenic Stroke and Underlying AF (CRYSTAL AF) trial, the 6-month rate was 8.9% in patients who had an implanted cardiac monitor versus 1.4% in patients who had routine evaluation. The lower rates of detection in CRYSTAL-AF compared with EMBRACE may have been attributable to a younger cohort of patients being evaluated; the mean age was 61.5 years in CRYSTAL-AF and 72.9 years in EMBRACE. In both trials, there was incremental detection of atrial fibrillation episodes over time, that is, new atrial fibrillation was identified continuously …